text
stringlengths
0
27.1M
meta
dict
# import aei import gdal import pickle import numpy as np import pandas as pd import matplotlib.pyplot as plt %matplotlib tk # set the working directories base = '/home/cba/Downloads/scale-conceptual/' plots = base + 'plots/' tch_file = base + 'marin_tch_byte.tif' gd_file = base + 'marin_tch_mask.tif' rnd_data = base + 'random-sampling-data.pck' # set flag for whether to calculate or load the random sampling calc_rs = False # set the number of resolutions to assess #res = [20, 30, 50, 100, 250, 500, 1000, 1250, 1500, 1750, 2000] res = [20, 30, 50, 100, 250, 500, 1000] # set the number of pixels to sample fullres = 2 * 2 testres = min(res) * min(res) n_var = testres / fullres n_loc = n_var n_rep = 1000 if calc_rs: # open the suitability raster and read in good indices gdref = gdal.Open(gd_file) gd = gdref.ReadAsArray() #gd_ext1 = gd == 1 gd_ext1 = np.where(gd == 1) bd = gd == 0 #gd_ext2 = np.where(gd > 1) gd = None gdref = None # read the tch data into memory tchref = gdal.Open(tch_file) #tchb1 = tchref.GetRasterBand(1) #ndval = tchb1.GetNoDataValue() tch = tchref.ReadAsArray().astype(np.float) tchb1 = None tchref = None # set the nodata values to nan tch[bd] = np.nan # create a set of arrays to store the data wg_var = np.zeros((len(res), n_rep, n_loc)) bg_var = np.zeros((len(res), n_rep, n_loc)) # loop through each resolution, and calculate within/between grain variance for i in range(len(res)): rdif = res[i] / 2 for j in range(n_rep): # find the random pixel locations to select loc_ext1 = np.random.choice(len(gd_ext1[0]), n_loc) #loc_ext2 = np.random.choice(len(gd_ext2[0]), n_loc) # loop through each random location for k in range(n_loc): # sample the x/y space around this pixel xmin = gd_ext1[0][loc_ext1[k]] - rdif xmax = gd_ext1[0][loc_ext1[k]] + rdif ymin = gd_ext1[1][loc_ext1[k]] - rdif ymax = gd_ext1[1][loc_ext1[k]] + rdif subset = tch[xmin:xmax, ymin:ymax] avg_ext1 = np.nanmean(subset) var_ext1 = np.nanvar(subset) wg_var[i,j,k] = var_ext1 bg_var[i,j,k] = avg_ext1 # then do it for the second extent #xmin = gd[0][loc_ext2[k]] - rdif #xmax = gd[0][loc_ext2[k]] + rdif #ymin = gd[1][loc_ext2[k]] - rdif #ymax = gd[1][loc_ext2[k]] + rdif #subset = tch[xmin:xmax, ymin:ymax] #avg_ext2 = np.nanmean(subset) #var_ext2 = np.nanvar(subset) # save this loop to a data file with open(rnd_data, 'w+') as f: pickle.dump([wg_var, bg_var], f) else: with open(rnd_data, 'r+') as f: wg_var, bg_var = pickle.load(f) # now that we have the data, reduce it to the plot the within/between grain variance wg_mean_loc = np.nanmean(wg_var, axis=2) # mean across all locations per rep wg_mean_all = np.nanmean(wg_mean_loc, axis=1) # mean of all locs, all reps wg_std = np.nanstd(wg_mean_loc, axis=1) # stdev of within grain variation across reps bg_var_loc = np.nanvar(bg_var, axis=2), # variance of between-grain measurements bg_mean = np.nanmean(bg_var_loc[0], axis=1) # mean of between-grain variance bg_std = np.nanstd(bg_var_loc[0], axis=1) # stdev of between-grain variance # plot these results cols = aei.color.color_blind(5) col = '#E3C16D' #col=cols[4] fill_alpha = 0.7 plt.figure(figsize=(4,3), dpi=200) # plot the standard deviations for each plt.fill_between(res, wg_mean_all-wg_std, wg_mean_all+wg_std, color=col, alpha=fill_alpha, label='Bootstrapped\nstandard deviation') plt.fill_between(res, bg_mean-bg_std, bg_mean+bg_std, color=col, alpha=fill_alpha) # set the line plots plt.plot(res, wg_mean_all, color='black', linewidth=1.5, linestyle = '-', label='Within-grain\nvariance') plt.plot(res, bg_mean, color='black', linewidth=1.5, linestyle = '--', label='Between-grain\nvariance') # set the labels plt.xlabel('Log grain size (m)') plt.ylabel('Spatial\nvariance (m{})'.format(r'$^2$')) plt.title('Scale-dependence in\nspatial tree height patterns') # log transform the axes #plt.yticks([], []) #ax = plt.gca() #ax.loglog() plt.xscale('log') #plt.yscale('log') # replace the axis labels #yticks=(20, 40, 60, 80, 100) yticks=[25, 50, 75, 100] #ylabels = ax.get_yticklabels() plt.yticks(yticks, ['{}'.format(f) for f in yticks]) plt.xticks(res, res) # set the custom legend lgd = plt.legend(loc='right', bbox_to_anchor=(1.7, 0.5), fancybox=True)#, # fancybox=True, shadow=True) # save the figure plt.tight_layout() plt.savefig('{}{}.png'.format(plots, 'Variance-plots'), dpi=300, additional_artists = [lgd], bbox_inches="tight") plt.savefig('{}{}.svg'.format(plots, 'Variance-plots'), additional_artists = [lgd], bbox_inches="tight")
{ "alphanum_fraction": 0.6294256491, "author": null, "avg_line_length": 33.4473684211, "converted": null, "ext": "py", "file": null, "hexsha": "f42ecba628c1621be6f014aa7951f72152e9670d", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2020-04-08T05:41:45.000Z", "max_forks_repo_forks_event_min_datetime": "2020-04-08T05:41:45.000Z", "max_forks_repo_head_hexsha": "ea5be566d810ca9d792625b2e97acd2065c9d09a", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "christobal54/aei-grad-school", "max_forks_repo_path": "projects/scale/scale-plot-spatial-variance.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "ea5be566d810ca9d792625b2e97acd2065c9d09a", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "christobal54/aei-grad-school", "max_issues_repo_path": "projects/scale/scale-plot-spatial-variance.py", "max_line_length": 85, "max_stars_count": null, "max_stars_repo_head_hexsha": "ea5be566d810ca9d792625b2e97acd2065c9d09a", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "christobal54/aei-grad-school", "max_stars_repo_path": "projects/scale/scale-plot-spatial-variance.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1543, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 5084 }
import pandas as pd import numpy as np #https://iq-inc.com/importerror-attempted-relative-import/ import data.data_retrieve as dr import features.data_preprocessing as pp import models.train_model as tm import models.train_model_FTR as tmFTR #could make it object oriented so that different models can be different objects #could make Preprocess Class data = dr.get_data() data = pp.feature_engineering(data) #data, X, y = pp.get_X_and_y_over_under(data,number_goals = 2.5) data, X, y = pp.get_X_and_y_FTR(data) X_train, y_train, X_test, y_test = pp.train_test_split(data,X,y) print(data.shape) print(X_train.shape, y_train.shape, X_test.shape, y_test.shape) #optimal_params = tm.get_optimal_parameters(X_train,y_train,n_trials=35) #print(optimal_params) optimal_params = { 'n_estimators': 286, 'max_depth': 6, 'num_leaves': 18, 'min_data_in_leaf': 70, 'feature_fraction': 0.4, 'lambda_l1': 10, 'lambda_l2': 55, 'bagging_fraction': 0.5, 'learning_rate': 0.03438248498061834, 'min_gain_to_split': 0.1030288398937825, 'bagging_freq': 1, } preds, preds_class = tmFTR.train_and_predict(X_train, y_train,X_test, y_test,optimal_params) evaluation_metrics = tmFTR.evaluate_predictions(y_test,preds, preds_class) cv_scores, shap_values, df_shaps= tm.get_cross_validated_scores_and_shap_values(X_train,y_train,optimal_params) print(df_shaps) #show performance of model on different teams (perhaps certain teams are predicted better)
{ "alphanum_fraction": 0.7020884521, "author": null, "avg_line_length": 33.2244897959, "converted": null, "ext": "py", "file": null, "hexsha": "3fd22b434f4f7aa3afeef1945640e691b4f570aa", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "e860b021d2e02d72f5223aee7ca4e33dacf4aa32", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "mkennealy/football-ML-models", "max_forks_repo_path": "src/main.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "e860b021d2e02d72f5223aee7ca4e33dacf4aa32", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "mkennealy/football-ML-models", "max_issues_repo_path": "src/main.py", "max_line_length": 111, "max_stars_count": null, "max_stars_repo_head_hexsha": "e860b021d2e02d72f5223aee7ca4e33dacf4aa32", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "mkennealy/football-ML-models", "max_stars_repo_path": "src/main.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 411, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1628 }
""" Provides a subclass to DSNFITSexaminer for plotting data. Possible symbols:: ================ =============================== character description ================ =============================== ``'o'`` circle marker ``'v'`` triangle_down marker ``'^'`` triangle_up marker ``'<'`` triangle_left marker ``'>'`` triangle_right marker ``'s'`` square marker ``'p'`` pentagon marker ``'D'`` diamond marker ``'h'`` hexagon1 marker ``'H'`` hexagon2 marker ``'1'`` tri_down marker ``'2'`` tri_up marker ``'3'`` tri_left marker ``'4'`` tri_right marker ``'*'`` star marker ``'+'`` plus marker ``'x'`` x marker ``'d'`` thin_diamond marker ``'|'`` vline marker ``'_'`` hline marker ``'.'`` point marker ``','`` pixel marker ``'-'`` solid line style ``'--'`` dashed line style ``'-.'`` dash-dot line style ``':'`` dotted line style ================ =============================== """ import logging import matplotlib as MPL import matplotlib.font_manager as MPLfont import numpy import os.path import pylab as PL import re import Data_Reduction.tipping as DRtip import Data_Reduction.FITS.SDFITSexaminer as FITSexam import support import support.graphics as GR logger = logging.getLogger(__name__) plotcolors = ['b','g','r','m','c'] plotsymbols = ['o','v','^','<','>', 's','p','D','h','H', '1','2','3','4','*', '+',"x","d","|","_"] fontP = MPLfont.FontProperties() fontP.set_size('xx-small') seconds_formatter = MPL.dates.DateFormatter("%H:%M:%S") class DSNFITSplotter(FITSexam.DSNFITSexaminer): """ """ def __init__(self, parent=None, FITSfile=None, hdulist=None): """ Create a new DSNFITSplotter object from an HDU list If invoked from within another object, then parent should be 'self'. Either a FITS file or an HDU list must be provided. """ mylogger = logging.getLogger(logger.name+".DSNFITSplotter") mylogger.debug("__init__: initializing superclass") FITSexam.DSNFITSexaminer.__init__(self, parent=parent, FITSfile=FITSfile, hdulist=hdulist) self.logger = mylogger self.plotter = {} for key in list(self.tables.keys()): table = self.tables[key] self.logger.debug("__init__: processing %s", table) self.plotter[key] = self.Plotter(self, table) self.logger.debug("__init__: completed") class Plotter(FITSexam.DSNFITSexaminer.Table): """ """ def __init__(self, parent, table): """ Initialization was already done for Table superclass """ self.logger = logging.getLogger(parent.logger.name+".Plotter") self.logger.debug("__init__: subclass of %s", table) for attr in list(table.__dict__.keys()): self.__setattr__(attr, table.__getattribute__(attr)) self.logger.debug("__init__: copying %s", attr) #------------------------ Plotter methods --------------------------------- def figure_rows_and_columns(self, naxes, **kwargs): """ Computes number of rows and columns for 'subplots' @param naxes : number of subplots (axes) needed @type naxes : int """ # get maximum number of rows and columns screensize = GR.get_screen_resolution() widthIN, heightIN = screensize['width']/100., screensize['height']/100. freewidth, freeheight = 0.95*widthIN, 0.95*heightIN self.logger.debug("figure_rows_and_columns: width available = %f in", freewidth) self.logger.debug("figure_rows_and_columns: height available = %f in", freeheight) if "width" in kwargs: width = kwargs["width"] # inches else: width = 4 if "heigth" in kwargs: height = kwargs["width"] # inches else: height = 4 max_cols = int(round(freewidth/width)) max_rows = int(round(freeheight/height)) self.logger.debug("figure_rows_and_columns: max cols and rows: %d,%d", max_cols, max_rows) max_graphs = max_cols*max_rows aspect = float(max_cols)/max_rows # how many figures do we need? num_figs = 1+naxes/max_graphs if num_figs > 1: num_cols = max_cols num_rows = max_rows else: num_cols = int(ceil(sqrt(aspect*naxes))) num_rows = int(ceil(float(naxes)/num_cols)) self.logger.debug("figure_rows_and_columns: %d rows, %d columns", num_rows, num_cols) return num_figs, num_rows, num_cols, (width,height) def init_multiplot(self, title, nrows, ncols, size=(4,4), sharey=False): """ create a figure with multiple plots sharing common X and Y axis The sublots have no space between them. When the number of graphs is large then multiple figures need to be created with 'init_multiplots' @param title : figure title @type title : str @param nrows : number of rows of subplots @type nrows : int @param ncols : number of columns of subplots @type ncols : int @param size : figure width, figure height (in) @type size : tuple of float @param sharey : same range on all Y axes (in) @type sharey : bool """ width, height = size fig, ax = PL.subplots(nrows=nrows, ncols=ncols, sharey=sharey) self.logger.debug("init_multiplot: %d rows, %d columns, size: %s", nrows, ncols, size) # could also be fig.set_size_inches(size, forward=True) fig.set_size_inches(ncols*width, nrows*height, forward=True) fig.subplots_adjust(wspace=0, hspace=0) # no space between plots in a row fig.suptitle(title) # adjust position to make room for figure legend. subplot_width = 0.8/ncols subplot_margin = 0.1*subplot_width subplot_offset = subplot_margin - 0.1 return fig, ax def init_multiplots(self, title, nfigs, nrows, ncols, size=(4,4), sharey=False): """ create multiple figures with multiple plots The sublots have no space between the axes in a figure. Use figure_rows_and_columns() to compute number of figures, rows, columns and figure size. @param title : figure title @type title : str @param nfigs : number of figures @type nfigs : int @param nrows : number of rows of subplots @type nrows : int @param ncols : number of columns of subplots @type ncols : int @param size : figure width, figure height (in) @type size : tuple of float @param sharey : same range on all Y axes (in) @type sharey : bool """ self.logger.debug("init_multiplots: %d figs, %d rows, %d columns", nfigs, nrows, ncols) figs = {} axs = {} for fignum in range(nfigs): figs[fignum], axs[fignum] = self.init_multiplot(title, nrows, ncols, size=size, sharey=sharey) return figs, axs def show_passband(self, figtitle=None, rows=None, savepath=None): """ Plots the passbands from this file as dynamic spectra If there are multiple beams, there will be a figure column for each beam and each pol. Else, if there is only one beam but multiple subchannels then there will be a figure column for each subchannel and each pol. Otherwise there is just a column for each pol. Image array structure --------------------- There is an image array for each subchannel, beam and pol combination. The data for each is a 2D nparray with dimensions (num_scans, num_chans). Initialization -------------- The spectra for each scan, subchannel, beam and pol combination are initialized as zeros with shape (32768,). The zeros are replaced with data from the FITS table DATA column. The images for each subchannel, beam and pol combination are initialized as numpy.arrays with shape (32768, 1). There is a flag dict 'start_image' which is initialized as True. When it is True, the initial image (see above) is replaced with data for the records in the first scan. The flag is then set to False. After data, the subimages for each scan are appended. The final image will have dimensions (num_scans*num_records, 32768). """ if rows == None: spectra = self.get_spectra(self.acs_rows) else: spectra = self.get_spectra(rows) # lay out the figure if self.props['num beams'] > 1: ncols = self.props['num beams']*self.props['num IFs'] elif self.props['num cycles'] > 1: ncols = self.props['num cycles']*self.props['num IFs'] else: ncols = self.props['num IFs'] # collect the diagnostic spectra if self.props["full Stokes"]: # when SPECTRA are Stokes parameters, plot IF power for two pol modes if self.props["num IFs"] == 2: datasource = "IFSPECTR" num_chans = self.props['num IFspec chans'] IFspectra = True else: if "SPECTRUM" in self.data.columns.names: datasource = "SPECTRUM" else: datasource = "DATA" IFspectra = False num_chans = self.props['num chans'] self.logger.debug("show_passband: data source is %s", datasource) # prepare empty images images = self.prepare_summary_images(num_chans) # get data statistics for scaling plots ymin, ymax, ymean, ystd = self.get_data_stats() # spectra are 2D arrays with frequency and scan as axes # images, also 2D arrays, are only needed if there is a TIME axis in # the data and then each record is a row in the array. cycles = self.cycle_keys # images with SCAN on the time axis have sub-images over record start_image = {} for cycle in cycles: subch_idx = cycle - 1 start_image[subch_idx] = {} for beam_idx in range(self.props["num beams"]): start_image[subch_idx][beam_idx] = {} for IF_idx in range(self.props["num IFs"]): start_image[subch_idx][beam_idx][IF_idx] = True # get the data for scan in self.scan_keys: scan_idx = self.scan_keys.index(scan) # scan numbers can start anywhere for cycle in cycles: subch_idx = cycle - 1 for beam_idx in range(self.props["num beams"]): beam = beam_idx+1 for IF_idx in range(self.props["num IFs"]): pol = IF_idx+1 #self.logger.debug("plot_passband: processing scan %d, subch %d, beam %d, pol %d", # scan, cycle, beam, pol) if self.props["time axis"]: # average scan spectrum and include record spectra in image # assume there is a scan number equal to the cycle number image, spectrum = \ self.average_records(scan, cycle, beam, pol) # this is the all-record sub-image for the scan if start_image[subch_idx][beam_idx][IF_idx]: images[subch_idx][beam_idx][IF_idx] = image start_image[subch_idx][beam_idx][IF_idx] = False else: images[subch_idx][beam_idx][IF_idx] = \ numpy.append(images[subch_idx][beam_idx][IF_idx], image, axis=0) else: # no time axis spec_indices = self.get_indices(scan=scan, cycle=cycle, beam=beam, pol=pol) image_line = self.data[datasource][spec_indices].reshape( num_chans,1).transpose() if start_image[subch_idx][beam_idx][IF_idx]: images[subch_idx][beam_idx][IF_idx] = image_line start_image[subch_idx][beam_idx][IF_idx] = False else: images[subch_idx][beam_idx][IF_idx] = \ numpy.append(images[subch_idx][beam_idx][IF_idx], image_line, axis=0) if figtitle: pass else: figtitle = os.path.basename(self.parent.file)[4:-5].replace('_',' ') fig, ax = self.init_multiplot(figtitle+" Spectrogram", nrows=1, ncols=ncols) # display the data # labels are updated only in the first time around a loop labels = {} num_beams = self.props['num beams'] num_cycles = len(self.cycle_keys) num_pols = self.props["num IFs"] halfband = self.get_first_good_value('BANDWIDT')/2.e6 for subch in range(num_cycles): for beam in range(self.props["num beams"]): for pol in range(self.props["num IFs"]): label = make_legend_labels(sckeys=list(range(num_cycles)), bmkeys=list(range(num_beams)), plkeys=list(range(num_pols)), sckey=subch, bmkey=beam, plkey=pol) if self.props['num beams'] > 1: col = 2*beam + pol elif self.props['num cycles'] > 1: col = 2*subch + pol else: col = pol # dynamic spectra of IF power ax[col].set_title(label) height, width = images[subch][beam][pol].shape ax[col].imshow(images[subch][beam][pol], aspect="auto", extent=(-halfband,halfband,0,height)) if col == 0: ax[col].set_ylabel("Cumulative record number") ax[col].set_xlabel("Frequency (MHz)") fig.subplots_adjust(top=0.88) fig.subplots_adjust(bottom=0.15) last_col = len(ax)-1 lines, labels = ax[last_col].get_legend_handles_labels() fig.legend(lines, labels, loc="upper right", ncol=2, prop = fontP) if savepath: fig.savefig(savepath) #fig.show() def show_all_spectra(self, rows=None, IFspectra=False, sharey=False): """ Plot spectra from a Table In each subplot are all the spectra for each beam and polarization from one row. If there are multiple records in a row, all records are plotted (not the average over records) @param rows : optional list of table rows; default: all @type rows : list of int @param IFspectra : plot IFSPECTR if there is one; default: False @type IFspectra : bool @param sharey : same range on all Y axes (in) @type sharey : bool """ # gets spectra from SPECTRUM column with RA and dec indices removed if rows == None: spectra = self.get_spectra(self.acs_rows) else: spectra = self.get_spectra(rows) num_graphs = len(spectra)/self.props['num cycles'] nfigs, nrows, ncols, size = self.figure_rows_and_columns(num_graphs) figs, axs = self.init_multiplots("Basic Spectra", nfigs, nrows, ncols, size, sharey=sharey) self.logger.debug("show_all_spectra: figure keys: %s", list(figs.keys())) self.logger.debug("show_all_spectra: axes keys: %s", list(axs.keys())) for fignum in list(figs.keys()): fig = figs[fignum] ax = axs[fignum] for row in range(nrows): for col in range(ncols): specidx = nrows*ncols*fignum + ncols*row + col if specidx >= len(spectra): break scan = self.data['SCAN'][specidx] cycle = self.data['CYCLE'][specidx] cycle_idx = self.cycle_keys.index(cycle) self.logger.debug( "show_all_spectra: doing row %d column %d spectrum %d", row, col, specidx) spec = spectra[specidx] # returns spectra from one row if self.props['full Stokes']: if "IFSPECTR" in self.data.columns.names and IFspectra: nchans = self.props['num IFspec chans'] npols = self.props['num IFs'] else: nchans = self.props['num chans'] npols = spec.shape[0] else: nchans = self.props['num chans'] npols = self.props['num IFs'] nbeams = self.props['num beams'] nrecs = self.props['num records'][cycle] self.logger.debug( "show_all_spectra: %d channels, %d pols, %d records", nchans, npols, nrecs) for rec in range(nrecs): record = rec+1 symbol = plotsymbols[rec % 20] for beam_idx in range(nbeams): beam = beam_idx+1 for pol_idx in range(npols): pol = pol_idx+1 # indices without row (0), RA (-1), dec (-2) if len(spec.shape) == 4: # with beam and time axis indices = self.get_indices(scan=scan, cycle=cycle, pol=pol, beam=beam, record=record, trimmed=True)[1:] elif len(spec.shape) == 3: # with time axis indices = self.get_indices(scan=scan, cycle=cycle, pol=pol, record=record, trimmed=True)[1:] elif len(spec.shape) == 2: indices = self.get_indices(scan=scan, cycle=cycle, pol=pol, trimmed=True)[1:] self.logger.debug("show_all_spectra: indices: %s", indices) color = plotcolors[beam_idx*npols+pol_idx % 5] trimmed = spec[indices] label = self.make_legend_labels(sckey=cycle_idx, bmkey=beam_idx, plkey=pol_idx) ax[row][col].plot(trimmed, color+',', label=label) # end loop over records ax[row][col].grid(True) ax[row][col].text(0.5, 0.95, 'scan '+str(scan), transform=ax[row][col].transAxes, horizontalalignment='center', verticalalignment='top') if row == nrows-1: for tick in ax[row][col].get_xticklabels(): tick.set_rotation(45) if col == 0: ax[row][col].set_ylabel("Power (counts)") # end loop over col # end loop over row lines, labels = ax[0][0].get_legend_handles_labels() fig.legend(lines, labels, loc="upper right", ncol=2, prop = fontP) # end loop over fig #fig.show() def make_legend_labels(self, dskey=None, tbkey=None, sckey=None, bmkey=None, plkey=None): dskeys=[] tbkeys=[] sckeys = list(range(self.props['num cycles'])) bmkeys = list(range(self.props['num beams'])) if self.props['full Stokes']: plkeys = list(range(4)) else: plkeys = list(range(self.props['num IFs'])) return make_legend_labels(dskeys=dskeys, tbkeys=tbkeys, sckeys=sckeys, bmkeys=bmkeys, plkeys=plkeys, dskey=dskey, tbkey=tbkey, sckey=sckey, bmkey=bmkey, plkey=plkey) def plot_BPSW_spectra(self, spectra=None, rows=None): """ plot reduced beam and position switched spectra @param spectra : optional dict of reduced spectra from 'BPSW_spectra()' @type spectra : numpy array @param rows : optional list of table rows to compute spectra; deault: all @type rows : list of int """ if spectra: npairs = len(spectra) elif rows : rows = self.acs_rows spectra = self.BPSW_spectra(rows) npairs = len(spectra) else: rows = self.acs_rows spectra = self.BPSW_spectra(rows) npairs = len(spectra) pairs = list(range(npairs)) nrows, ncols = self.figure_rows_and_columns(len(spectra)) fig, ax = self.init_multiplot("BPSW Spectra", nrows, ncols) for row in range(nrows): for col in range(ncols): pair = row*ncols + col spec = spectra[pair] nrecs, npols, nchans = spec.shape for rec in range(nrecs): symbol = plotsymbols[rec % 20] for pol in range(npols): color = plotcolors[pol % 5] trimmed = support.trim_extremes(spec[rec, pol]) ax[row][col].plot(trimmed, color+',', label="rec"+str(rec)+" P"+str(pol+1)) ax[row][col].grid(True) ax[row][col].text(0.5, 0.95, 'pair '+str(pair+1), transform=ax[row][col].transAxes, horizontalalignment='center', verticalalignment='top') if row == nrows-1: for tick in ax[row][col].get_xticklabels(): tick.set_rotation(45) if col == 0: ax[row][col].set_ylabel("Normalized Power") lines, labels = ax[0][0].get_legend_handles_labels() fig.legend(lines, labels, loc="upper right", ncol=2, prop = fontP) show() def plot_PSSW_spectra(self, figtitle=None, scans=[], savepath=None): """ Plot position switched spectra We need to check self.data['OBSMODE'] @param scans : list of scan numbers, ons and offs @type scans : list of int """ if scans == []: scans = self.scan_keys # lay out the figure if self.props['num beams'] > 1: ncols = self.props['num beams']*self.props['num IFs'] elif self.props['num cycles'] > 1: ncols = self.props['num cycles']*self.props['num IFs'] else: ncols = self.props['num IFs'] if figtitle: pass else: first = os.path.basename(self.parent.file).index('_')+1 figtitle = "DSS-"+str(self.dss) +" "+ \ os.path.basename(self.parent.file)[first:-5] fig, ax = self.init_multiplot(figtitle+" (Sig-Ref)/Ref", nrows=1, ncols=ncols) # data source if 'IFSPECTR' in self.data.names: # this is for Stokes spectra datasource = 'IFSPECTR' elif 'SPECTRUM' in self.data.names: datasource = 'SPECTRUM' else: datasource = 'DATA' self.logger.debug("plot_PSSW_spectra: data column is %s", datasource) labels = {} num_beams = self.props['num beams'] num_cycles = len(self.cycle_keys) num_pols = self.props["num IFs"] for subch_idx in range(num_cycles): cycle = subch_idx + 1 for beam_idx in range(self.props["num beams"]): beam = beam_idx + 1 for pol_idx in range(self.props["num IFs"]): pol = pol_idx+1 label = make_legend_labels(sckeys=list(range(num_cycles)), bmkeys=list(range(num_beams)), plkeys=list(range(num_pols)), sckey=subch_idx, bmkey=beam_idx, plkey=pol_idx) if self.props['num beams'] > 1: col = 2*beam_idx + pol_idx elif self.props['num cycles'] > 1: col = 2*subch_idx + pol_idx else: col = pol_idx # dynamic spectra of IF power ax[col].set_title(label) self.logger.debug("plot_PSSW_spectra: subch %d beam %d pol %d", cycle, beam, pol) #on_scans = numpy.unique(self.data['SCAN'][numpy.where( # self.data['SIG'][self.acs_rows] == True)[0]]) #of_scans = numpy.unique(self.data['SCAN'][numpy.where( # self.data['SIG'][self.acs_rows] == False)[0]]) on_indices = numpy.where(self.data['SIG'] == True)[0] of_indices = numpy.where(self.data['SIG'] == False)[0] on_scans = numpy.unique(self.data['SCAN'][numpy.intersect1d(on_indices, self.rows)]) of_scans = numpy.unique(self.data['SCAN'][numpy.intersect1d(of_indices, self.rows)]) num_pairs = min(len(on_scans), len(of_scans)) for index in range(num_pairs): on_scan = on_scans[index] of_scan = of_scans[index] # get the indices for the DATA cell in the ON position try: indices = self.get_indices(scan=on_scan, cycle=cycle, pol=pol, beam=beam) self.logger.debug("plot_PSSW_spectra: scan %d on indices: %s", on_scan, indices) except ValueError as details: self.logger.warning("plot_PSSW_spectra: %s", str(details)) continue row = indices[0] # get the X-axis units for the ON position if datasource == 'IFSPECTR': v = self.compute_X_axis(row, frame='DELF-OBS', num_chans=self.props['num IFspec chans']) else: v = self.compute_X_axis(row, frame='DELF-OBS', num_chans=self.props['num chans']) on = self.data[datasource][indices] # get the indices for the DATA cell in the OFF position try: ref_indices = self.get_indices(scan=of_scan, cycle=cycle, pol=pol, beam=beam) self.logger.debug("plot_PSSW_spectra: scan %d off indices: %s", of_scan, ref_indices) off = self.data[datasource][ref_indices] except ValueError as details: self.logger.warning("plot_PSSW_spectra: %s", str(details)) continue spectrum = (on-off)/off label = "("+str(on_scan)+"-"+str(of_scan)+")/"+str(of_scan) ax[col].plot(v, spectrum, label=label) if index == 0: ax[col].grid() ax[col].set_xlabel("Frequency (MHz)") #heading = "%8.3f MHz" % (self.data['OBSFREQ'][row]/1e6) #heading += " P"+str(pol) #ax[col].set_title(heading) last_col = len(ax)-1 lines, labels = ax[last_col].get_legend_handles_labels() fig.legend(lines, labels, loc="upper right", ncol=2, prop = fontP) if savepath: savefig(savepath) show() def plot_line(self, rows=[], window=(-100,100), frame="RADI-OBJ", source='67P', savepath=None): """ plot reduced averaged spectral lines (from 'reduce_line()) for both pols @param rows : optional rows to include; default: all @type rows : list of int @param window : optional left and right limits of the X axis (-100,100) @type window : tuple of float @param frame : optional reference frame for Doppler calc ("RADI-OBJ") @type frame : str @param source : source name for Doppler calculation; default: 67P @type source : str @param savepath : optional path for saved image; default: no save @type savepath : str """ try: x, y, rms, Tsys, intgr = self.reduce_line(rows=rows, window=window, frame=frame, source=source) except TypeError: self.logger.error("plot_line: nothing to plot") else: figure() plot(x, y[0], label="Pol1") plot(x, y[1], label="Pol2") grid() xlim(*window) legend(prop=fontP) titlestr = \ source+" DSS-"+str(self.dss)+" "+self.datestr+" "+self.timestr title(titlestr) xlabel("$V_{LSR}$ (km s$^{-1}$") ylabel("$T_{ant}$ (K)") if savepath: fname = titlestr.replace("/","-").replace(" ","_")+".png" savefig(savepath+fname) return {"rms": rms, "Tsys": Tsys, "intgr": intgr} def plot_all_Tsys(self): """ Displays all Tsys values so user can select row range This works for WVSR data but needs to be elaborated for SAO data """ figure() stride = self.props['num cycles']*self.props['num IFs'] for cycle_idx in range(self.props['num cycles']): for beam_idx in range(self.props['num beams']): for pol_idx in range(self.props['num IFs']): label = self.make_legend_labels(sckey=cycle_idx, bmkey=beam_idx, plkey=pol_idx) for sig in [True, False]: if sig: lbl = label+" on" ls = "-" index = cycle_idx else: lbl = label+" off" ls = "--" index = 2+cycle_idx plot(self.data['TSYS'][index:len(self.acs_rows):stride, pol_idx,0,0,0], ls=ls, label=lbl) grid() legend(loc='best') xlabel("row") ylabel("average power") title(self.datestr) #show() def plot_Tsys(self, good_wx_data=None, X=None): """ Plot average power versus time or airmass or list index Options for X are:: "time" - time of measurement "airmass" - 1/sin(elev) None - list index """ if good_wx_data: pass else: good_wx_data = self.get_wx_datacubes() heading = "DSS-%2d %s" % (self.dss, self.datestr) fig, ax = self.init_multiplot(heading, 1, self.props['num cycles']) for subch in self.cycle_keys: cycle_idx = subch - 1 for beam in range(self.props['num beams']): for pol in range(self.props['num IFs']): for sig in [True, False]: tm = good_wx_data['UNIXtime'][sig] plottimes = MPL.dates.epoch2num(tm) tsys = good_wx_data['TSYS'][sig][:,cycle_idx,beam,pol] el = good_wx_data['ELEVATIO'][sig] label = self.make_legend_labels(bmkey=beam, plkey=pol) color_idx = 2*int(sig) + pol if sig: lbl = label+" sig" ls = "-" else: lbl = label+" ref" ls = "--" if X == "time": ax[cycle_idx].plot_date(plottimes, tsys, linestyle=ls, marker='.', color=plotcolors[color_idx], label=lbl) elif X == "airmass": ax[cycle_idx].plot(1/sin(pi*array(el)/180.), tsys, color=plotcolors[color_idx], marker='.', ls=ls, label=lbl) else: ax[cycle_idx].plot(tsys, color=plotcolors[color_idx], marker='.', ls=ls, label=lbl) ax[cycle_idx].grid(True) ax[cycle_idx].legend(loc='best', fontsize='xx-small', numpoints=1) if X == "time": ax[cycle_idx].xaxis.set_major_formatter(seconds_formatter) fig.autofmt_xdate() elif X == "airmass": ax[cycle_idx].set_xlabel("Airmass") else: ax[cycle_idx].set_xlabel("index") ax[cycle_idx].set_title("subchannel "+str(subch)) #fig.show() #--------------------------- DSNFITSplotter methods ------------------------- def plot_average(self, frame='RADI-LSR', source=None, xlimits=None, ylimits=None): """ plot an averaged DSN FITS spectrum A DSN FITS spectrum is multi-dimensional with axes:: [[beam,] [time,]], pol, [dec, RA], frequency-like [] indicates the axes may not be present. During data manipulations the [ra, dec,] axes are the first to be eliminated. Note that the frame of the provided spectrum is not changed. Only the X-axis is recomputed before plotting. @param row : row to be used to get observing parameters @type row : int @param frame : frame in which to plot the data @type frame : str @param xlimits : minimum and maximum X-axis values @type xlimits : (float, float) @param ylimits : minimum and maximum Y-axis values @type ylimits : (float, float) @param average : combine the two polarizations @type average : bool """ if type(spectrum) == FITSexam.DSNFITSexaminer.Table.Data: # this is a window into the original spectrum x = spectrum.x y = spectrum.y frame = spectrum.frame # change frame if desired if frame == spectrum.frame: self.logger.debug("plot_spectrum: keeping frame %s", frame) else: # change to the requested frame self.logger.debug("plot_spectrum: changing to frame %s", frame) new_x = spectrum.compute_X_axis(row, frame) x = new_x[window.channels[0]:window.channels[1]] self.logger.debug("plot_spectrum: plotting in frame %s", self.frame) else: # get SPECTRUM from current row self.logger.info("plot_spectrum: data is not a class Data instance") spectrum = self.get_spectra(row) if xlimits: ch1, ch2 = xlimits[0], xlimits[1] if frame == "RADI-OBJ" and self.frame != "RADI-OBJ": # Change to the object's rest frame source = self.dataset[row]["OBJECT"] if re.match('67P', source): self.logger.debug("plot_spectrum: using 67P frame") vspline = self.load_ephemeris('67P') x = spectrum.compute_X_axis(row=row, frame="RADI-OBJ", vspline=vspline)[ch1:ch2] y = spectrum[ch1:ch2] else: self.logger.error("plot_spectrum: no ephemeris for %s", self.object) raise RuntimeException("cannot compute velocity of %s" % source) else: x = spectrum.compute_X_axis(row=row)[ch1:ch2] y = spectrum[ch1:ch2] self.logger.debug("plot_spectrum: x = %s", x) self.logger.debug("plot_spectrum: y = %s", y) # make the plot figure() rms = {} for pol in [0,1]: plot(x, y[pol], label="pol "+str(pol)+" ("+("%d"%self.tau[pol])+"s)") rms[pol] = y[pol].std() if average: ave = (y[0]+y[1])/2 plot(x, ave, label="both ("+("%d"%(self.tau[0]+self.tau[1]))+"s)") rms['avg'] = ave.std() if xlimits: xlim(*xlimits) if ylimits: ylim(*ylimits) if self.frame == "CHAN-OBS": xlabel("Channel") elif self.frame[:5] == ("OPTI-" or "RADI-" or "RELA-"): if self.frame[4:] == "-OBS": xlabel(r"$V_{obs} (\mbox{km s}^{-1})$") else: xlabel("Frequency (MHz)") ylabel(r"Antenna Temperature (K)") grid() legend() titlestr = support.text.clean_TeX(str(ds0.year)+"/"+str(ds0.doy)) title(titlestr) show() self.logger.info("plot_spectrum: pol0, pol1, avg: %s", rms) return rms #------------------- class for plotting TIPPING CURVE extension data ---------- class TidTipPlotter(FITSexam.TidTipAnalyzer): """ class for plotting data from the TIPPING CURVE extensions """ def __init__(self, extension): """ """ FITSexam.TidTipAnalyzer.__init__(self, extension) def plot_data(): """ """ fig = figure() for IF in range(4): PM = IF+1 rows = where(self.data['CYCLE'] == PM) tsys = self.data['TSYS'][rows] am = DRtip.airmass(self.data['ELEVATIO'][rows]) plot(am, tsys, plotcolors[IF]+'-') plot(am, tsys, plotcolors[IF]+'.', label="IF%d" % PM) # add fit to legend label = "IF%d %5.1f$\pm$%3.1f %5.3f$\pm$%5.3f" % ( IF, Trx[IF],sigTrx[IF], tau[IF],sigtau[IF]) handles[IF].set_label(label) legend(loc="upper left", numpoints=1) grid() title(self.header['DATE-OBS']) xlabel("airmass") legend(loc="upper left", numpoints=1) show() if project: sessiondir = projects_dir+project+"/Observations/dss43/%4d/%03d/" % ( self.year, self.DOY) fig.savefig(sessiondir+"tipdatafit-"+str(index+1)+".png") #----------------------------------- module functions ------------------------- def make_legend_labels(dskeys=[], tbkeys=[], sckeys=[], bmkeys=[], plkeys=[], dskey=None, tbkey=None, sckey=None, bmkey=None, plkey=None): """ @param dskeys : all datafile or examiner keys @param tbkeys : all table keys @param sckeys : all subchannel keys @param bmkeys : all beam keys @param plkeys : all polarization keys @param dskey : datafile or examiner key @param tbkey : table key @param sckey : subchannel key @param bmkey : beam key @param plkeys :polarization key """ label = "" if dskey != None and len(dskeys) > 1: label += "ds"+str(dskey+1) if tbkey != None and len(tbkeys) > 1: label += " tb"+str(tbkey+1) if sckey != None and len(sckeys) > 1: label += " sc"+str(sckey+1) if bmkey != None and len(bmkeys) > 1: label += " B"+str(bmkey+1) if plkey != None and len(plkeys) > 1: label += "P"+str(plkey+1) return label def get_power_range(table, subch, beam, pol): """ calculate limits for power levels the idea here is to avoid huge spikes compressing the spectra will need some tuning @param subch : sub-channel number @param beam : beam number (1-based) @param pol : polarization number (not NRAO pol code) """ subch_idx = subch-1 beam_idx = beam-1 IF_idx = pol-1 # gets the indices for scan 1 indices = table.get_indices(cycle=subch, beam=beam, IF=pol) logger.debug("get_power_range: indices: %s", indices) # gets spectra from all rows spectrum = table.data['SPECTRUM'][:,indices[1:]] logger.debug("get_power_range: spectrum shape is %s", spectrum.shape) mean_pwr = spectrum.mean() pwr_std = spectrum.std() max_pwr = spectrum.max() min_pwr = spectrum.min() if max_pwr > mean_pwr + 4*pwr_std: ymax = mean_pwr + pwr_std else: ymax = mean_pwr + 4*pwr_std if min_pwr < mean_pwr - 4*pwr_std: ymin = mean_pwr - pwr_std else: ymin = mean_pwr - 4*pwr_std return ymin, ymax
{ "alphanum_fraction": 0.5367081632, "author": null, "avg_line_length": 40.4488348531, "converted": null, "ext": "py", "file": null, "hexsha": "6facd4eefdba2e2ac4c0df75b4116e52909233dd", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "f007d716b5c28c086910a81206cffaf37ff6368c", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "SDRAST/Data_Reduction", "max_forks_repo_path": "FITS/SDFITSplotter.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "f007d716b5c28c086910a81206cffaf37ff6368c", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "SDRAST/Data_Reduction", "max_issues_repo_path": "FITS/SDFITSplotter.py", "max_line_length": 96, "max_stars_count": null, "max_stars_repo_head_hexsha": "f007d716b5c28c086910a81206cffaf37ff6368c", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "SDRAST/Data_Reduction", "max_stars_repo_path": "FITS/SDFITSplotter.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 9638, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 39923 }
import argparse import os import pandas as pd import imagesize import nltk from nltk.corpus import wordnet as wn import numpy as np import re from ..create.scads_classes import Node, Image, Clip from ..create.create_scads import add_conceptnet from ..create.add_datasets import add_dataset from .wnids_to_concept import SYNSET_TO_CONCEPTNET_ID class DatasetInstaller: def get_name(self): raise NotImplementedError() def get_data(self, dataset, session, root): raise NotImplementedError() def get_conceptnet_id(self, label): return "/c/en/" + label.lower().replace(" ", "_").replace("-", "_") class ImageClassificationInstaller(DatasetInstaller): def get_name(self): raise NotImplementedError() def get_data(self, dataset, session, root): size = "full" modes = ['train', 'test'] label_to_node_id = {} all_images = [] for mode in modes: df_label = pd.read_feather( os.path.join(root, dataset.path, "labels_" + size, 'labels_' + mode + '.feather')) df = pd.crosstab(df_label['id'], df_label['class']) mode_dir = os.path.join(dataset.path, f'{dataset.path}_' + size, mode) for image in os.listdir(os.path.join(root, mode_dir)): if image.startswith('.'): continue label = df.loc[image].idxmax() # Get node_id if label in label_to_node_id: node_id = label_to_node_id[label] else: node = session.query(Node).filter_by(conceptnet_id=self.get_conceptnet_id(label)).first() node_id = node.id if node else None label_to_node_id[label] = node_id # Scads is missing a missing conceptnet id if node_id is None: continue img = Image(dataset_id=dataset.id, node_id=node_id, path=os.path.join(mode_dir, image)) all_images.append(img) return all_images class ObjectDetectionInstaller(DatasetInstaller): def get_name(self): raise NotImplementedError() def get_data(self, dataset, session, root): size = "full" modes = ['train', 'test'] label_to_node_id = {} all_images = [] for mode in modes: df_label = pd.read_feather( os.path.join(root, dataset.path, "labels_" + size, 'labels_' + mode + '.feather')) bbox = list(df_label.loc[:, 'bbox'].copy()) for i in range(len(bbox)): bbx = bbox[i].split(',') x_min = float(bbx[0].strip()) y_min = float(bbx[1].strip()) x_max = float(bbx[2].strip()) y_max = float(bbx[3].strip()) w = x_max - x_min h = y_max - y_min area = w * h bbox[i] = area df_label.loc[:, 'bbox'] = bbox pt = df_label.pivot_table(index='id', columns='class', values='bbox', aggfunc=np.max) mode_dir = os.path.join(dataset.path, f'{dataset.path}_' + size, mode) for image in os.listdir(os.path.join(root, mode_dir)): if image.startswith('.'): continue width, height = imagesize.get(os.path.join(root, mode_dir) + '/' + image) img_size = width * height label = pt.loc[image].dropna().idxmax() bbox_area = pt.loc[image, label] if bbox_area <= img_size * .2: continue # Get node_id if label in label_to_node_id: node_id = label_to_node_id[label] else: node = session.query(Node).filter_by(conceptnet_id=self.get_conceptnet_id(label)).first() node_id = node.id if node else None label_to_node_id[label] = node_id # Scads is missing a missing conceptnet id if not node_id: continue img = Image(dataset_id=dataset.id, node_id=node_id, path=os.path.join(mode_dir, image)) all_images.append(img) return all_images class VideoClassificationInstaller(DatasetInstaller): def get_name(self): raise NotImplementedError() def composed_labels(self, string, dataset): if dataset.name == 'HMDB': return [w.strip() for w in string.split('_')] elif dataset.name == 'UCF101': list_words = re.findall('[A-Z][^A-Z]*', string) return [w.lower().strip() for w in list_words] def get_data(self, dataset, session, root): size = "full" modes = ['train', 'test'] label_to_node_id = {} all_clips = [] for mode in modes: base_path = os.path.join(dataset.path, dataset.path + '_' + size, mode) #print(base_path, dataset.path, root) df = pd.read_feather( os.path.join(root, dataset.path, "labels_" + size, 'labels_' + mode + '.feather')) if mode == "test": df_label = pd.crosstab(df['id'], df['class']) df = pd.read_feather( os.path.join(root, dataset.path, "labels_" + size, "meta_" + mode + ".feather") ) for _, row in df.iterrows(): row = row.astype("object") if mode == "test": label = df_label.loc[row['id']].idxmax() else: label = row['class'] if label in label_to_node_id: node_id = label_to_node_id[label] clip = Clip( clip_id=row['id'], video_id=row['video_id'], base_path=base_path, start_frame=row['start_frame'], end_frame=row['end_frame'], real_label=self.get_conceptnet_id(label).split('/')[-1], dataset_id=dataset.id, node_id=node_id ) all_clips.append(clip) else: node = session.query(Node).filter_by(conceptnet_id=self.get_conceptnet_id(label)).first() # If the node related to the class doesn't exist if node: node_id = node.id label_to_node_id[label] = node_id clip = Clip( clip_id=row['id'], video_id=row['video_id'], base_path=base_path, start_frame=row['start_frame'], end_frame=row['end_frame'], real_label=self.get_conceptnet_id(label).split('/')[-1], dataset_id=dataset.id, node_id=node_id ) all_clips.append(clip) # Else, we decompose the class name and check for concepts corresponding to each word else: labels = self.composed_labels(label, dataset) for l in labels: if l in label_to_node_id: node_id = label_to_node_id[l] else: node = session.query(Node).filter_by(conceptnet_id=self.get_conceptnet_id(l)).first() node_id = node.id if node else None label_to_node_id[l] = node_id # Handle the case when a classe, even decomposed, is not assign to any concept if not node_id: continue real = '_'.join(labels) clip = Clip( clip_id=row['id'], video_id=row['video_id'], base_path=base_path, start_frame=row['start_frame'], end_frame=row['end_frame'], real_label=self.get_conceptnet_id(real).split('/')[-1], dataset_id=dataset.id, node_id=node_id ) all_clips.append(clip) return all_clips class CifarInstallation(ImageClassificationInstaller): def get_name(self): return "CIFAR100" class MnistInstallation(ImageClassificationInstaller): def get_name(self): return "MNIST" def get_conceptnet_id(self, label): mnist_classes = { '0': '/c/en/zero', '1': '/c/en/one', '2': '/c/en/two', '3': '/c/en/three', '4': '/c/en/four', '5': '/c/en/five', '6': '/c/en/six', '7': '/c/en/seven', '8': '/c/en/eight', '9': '/c/en/nine', } return mnist_classes[label] class ImageNetInstallation(ImageClassificationInstaller): def get_name(self): return "ImageNet" def get_conceptnet_id(self, label): return SYNSET_TO_CONCEPTNET_ID[label] class ImageNet22kInstallation(ImageClassificationInstaller): def wnid_to_name(self, wnid): synset = wn.synset_from_pos_and_offset('n', int(wnid[1:])) return synset.lemmas()[0].name() def get_name(self): return "ImageNet22k" def get_data(self, dataset, session, root): nltk.download('wordnet') all_images = [] all_wnids = os.listdir(os.path.join(root, dataset.path)) for wnid in all_wnids: label = self.wnid_to_name(wnid) node = session.query(Node).filter_by(conceptnet_id=self.get_conceptnet_id(label)).first() node_id = node.id if node else None if not node_id: continue for image in os.listdir(os.path.join(root, dataset.path, wnid)): img = Image(dataset_id=dataset.id, node_id=node_id, path=os.path.join(dataset.path, wnid, image)) all_images.append(img) return all_images def get_conceptnet_id(self, label): if label in SYNSET_TO_CONCEPTNET_ID: return SYNSET_TO_CONCEPTNET_ID[label] else: return super().get_conceptnet_id(label) class COCO2014Installation(ObjectDetectionInstaller): def get_name(self): return "COCO2014" def get_conceptnet_id(self, label): label_to_label = {1: 'person', 2: 'bicycle', 3: 'car', 4: 'motorcycle', 5: 'airplane', 6: 'bus', 7: 'train', 8: 'truck', 9: 'boat', 10: 'traffic light', 11: 'fire hydrant', 12: '-', 13: 'stop sign', 14: 'parking meter', 15: 'bench', 16: 'bird', 17: 'cat', 18: 'dog', 19: 'horse', 20: 'sheep', 21: 'cow', 22: 'elephant', 23: 'bear', 24: 'zebra', 25: 'giraffe', 26: '-', 27: 'backpack', 28: 'umbrella', 29: '-', 30: '-', 31: 'handbag', 32: 'tie', 33: 'suitcase', 34: 'frisbee', 35: 'skis', 36: 'snowboard', 37: 'sports ball', 38: 'kite', 39: 'baseball bat', 40: 'baseball glove', 41: 'skateboard', 42: 'surfboard', 43: 'tennis racket', 44: 'bottle', 45: '-', 46: 'wine glass', 47: 'cup', 48: 'fork', 49: 'knife', 50: 'spoon', 51: 'bowl', 52: 'banana', 53: 'apple', 54: 'sandwich', 55: 'orange', 56: 'broccoli', 57: 'carrot', 58: 'hot dog', 59: 'pizza', 60: 'donut', 61: 'cake', 62: 'chair', 63: 'couch', 64: 'potted plant', 65: 'bed', 66: '-', 67: 'dining table', 68: '-', 69: '-', 70: 'toilet', 71: '-', 72: 'tv', 73: 'laptop', 74: 'mouse', 75: 'remote', 76: 'keyboard', 77: 'cell phone', 78: 'microwave', 79: 'oven', 80: 'toaster', 81: 'sink', 82: 'refrigerator', 83: '-', 84: 'book', 85: 'clock', 86: 'vase', 87: 'scissors', 88: 'teddy bear', 89: 'hair drier', 90: 'toothbrush', 91: '-', 92: ''} return "/c/en/" + label_to_label[label].lower().replace(" ", "_").replace("-", "_") class DomainNetInstallation(ImageClassificationInstaller): def __init__(self, domain_name): self.domain = domain_name def get_name(self): return "DomainNet: " + self.domain def get_conceptnet_id(self, label): exceptions = {'paint_can': 'can_of_paint', 'The_Eiffel_Tower': 'eiffel_tower', 'animal_migration': 'migration', 'teddy-bear': 'teddy_bear', 'The_Mona_Lisa': 'mona_lisa', 't-shirt': 't_shirt', 'The_Great_Wall_of_China': 'great_wall_of_china'} if label in exceptions: return "/c/en/" + exceptions[label] return "/c/en/" + label.lower().replace(" ", "_").replace("-", "_") class MslCuriosityInstallation(ImageClassificationInstaller): def get_name(self): return "MslCuriosity" def get_conceptnet_id(self, label): exceptions = {'drill holes' : 'holes', 'observation tray' : 'tray', 'rover rear deck' : 'deck'} if label in exceptions: return "/c/en/" + exceptions[label] return "/c/en/" + label.lower().replace(" ", "_").replace("-", "_") class MarsSurfaceInstallation(ImageClassificationInstaller): def get_name(self): return "MarsSurface" def get_conceptnet_id(self, label): exceptions = {'drill holes' : 'holes', 'observation tray' : 'tray', 'rover rear deck' : 'deck'} if label in exceptions: return "/c/en/" + exceptions[label] return "/c/en/" + label.lower().replace(" ", "_").replace("-", "_") class VOC2009Installation(ObjectDetectionInstaller): def get_name(self): return "VOC2009" def get_conceptnet_id(self, label): exceptions = {'pottedplant': 'potted_plant', 'tvmonitor': 'tv_monitor', 'diningtable': 'dining_table'} if label in exceptions: return "/c/en/" + exceptions[label] return "/c/en/" + label.lower().replace(" ", "_").replace("-", "_") class GoogleOpenImageInstallation(ObjectDetectionInstaller): def get_name(self): return "GoogleOpenImage" class HMDBInstallation(VideoClassificationInstaller): def get_name(self): return "HMDB" class UCF101Installation(VideoClassificationInstaller): def get_name(self): return "UCF101" def get_conceptnet_id(self, label): exceptions = {'Skijet': 'jet_ski'} if label in exceptions: return "/c/en/" + exceptions[label] else: label_clean = '_'.join([i.lower() for i in re.findall('[A-Z][^A-Z]*', label)]) if len(label_clean) != 0: return "/c/en/" + label_clean#label.lower().replace(" ", "_")#.replace("-", "_") else: return "/c/en/" + label.lower() class Installer: def __init__(self, path_to_database): self.db = path_to_database def install_conceptnet(self, path_to_conceptnet): add_conceptnet(self.db, path_to_conceptnet) def install_dataset(self, root, path_to_dataset, dataset_installer): add_dataset(self.db, root, path_to_dataset, dataset_installer) if __name__ == "__main__": # Get arguments parser = argparse.ArgumentParser(description='Scads') parser.add_argument("--db", type=str, help="Path to database", required=True) parser.add_argument("--conceptnet", type=str, help="Path to ConceptNet directory") parser.add_argument("--root", type=str, help="Root containing dataset directories") parser.add_argument("--cifar100", type=str, help="Path to CIFAR100 directory from the root") parser.add_argument("--mnist", type=str, help="Path to MNIST directory from the root") parser.add_argument("--imagenet", type=str, help="Path to ImageNet directory from the root") parser.add_argument("--imagenet22k", type=str, help="Path to ImageNet22k directory from the root") parser.add_argument("--coco2014", type=str, help="Path to COCO2014 directory from the root") parser.add_argument("--voc2009", type=str, help="Path to voc2009 directory from the root") parser.add_argument("--googleopenimage", type=str, help="Path to googleopenimage directory from the root") parser.add_argument("--domainnet", nargs="+") parser.add_argument("--hmdb", type=str, help="Path to hmdb directory from the root") parser.add_argument("--ucf101", type=str, help="Path to ufc101 directory from the root") parser.add_argument("--msl_curiosity", type=str, help="Path to msl_curiosity directory from the root") parser.add_argument("--mars_surface_imgs", type=str, help="Path to mars_surface_imgs directory from the root") args = parser.parse_args() # Install SCADS installer = Installer(args.db) if args.conceptnet: installer.install_conceptnet(args.conceptnet) if not args.root: raise RuntimeError("Must specify root directory.") if args.cifar100: installer.install_dataset(args.root, args.cifar100, CifarInstallation()) if args.mnist: installer.install_dataset(args.root, args.mnist, MnistInstallation()) if args.imagenet: installer.install_dataset(args.root, args.imagenet, ImageNetInstallation()) if args.imagenet22k: installer.install_dataset(args.root, args.imagenet22k, ImageNet22kInstallation()) if args.coco2014: installer.install_dataset(args.root, args.coco2014, COCO2014Installation()) if args.voc2009: installer.install_dataset(args.root, args.voc2009, VOC2009Installation()) if args.googleopenimage: installer.install_dataset(args.root, args.googleopenimage, GoogleOpenImageInstallation()) if args.domainnet: for domain in args.domainnet: name = domain.split("-")[1].capitalize() installer.install_dataset(args.root, domain, DomainNetInstallation(name)) if args.hmdb: installer.install_dataset(args.root, args.hmdb, HMDBInstallation()) if args.ucf101: if not args.root: raise RuntimeError("Must specify root directory.") installer.install_dataset(args.root, args.ucf101, UCF101Installation()) if args.msl_curiosity: if not args.root: raise RuntimeError("Must specify root directory.") installer.install_dataset(args.root, args.msl_curiosity, MslCuriosityInstallation()) if args.mars_surface_imgs: if not args.root: raise RuntimeError("Must specify root directory.") installer.install_dataset(args.root, args.mars_surface_imgs, MarsSurfaceInstallation())
{ "alphanum_fraction": 0.5398881218, "author": null, "avg_line_length": 42.5815450644, "converted": null, "ext": "py", "file": null, "hexsha": "7e161ca862e61e36df6c1b3e4d05470e87a9edb5", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 2, "max_forks_repo_forks_event_max_datetime": "2022-02-27T04:27:48.000Z", "max_forks_repo_forks_event_min_datetime": "2022-02-14T22:40:29.000Z", "max_forks_repo_head_hexsha": "0fa9ebeccc9177069aa09b2da84746b7532e3495", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "BatsResearch/taglets", "max_forks_repo_path": "taglets/scads/create/install.py", "max_issues_count": 1, "max_issues_repo_head_hexsha": "0fa9ebeccc9177069aa09b2da84746b7532e3495", "max_issues_repo_issues_event_max_datetime": "2021-11-10T16:01:47.000Z", "max_issues_repo_issues_event_min_datetime": "2021-11-10T16:01:47.000Z", "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "BatsResearch/taglets", "max_issues_repo_path": "taglets/scads/create/install.py", "max_line_length": 119, "max_stars_count": 13, "max_stars_repo_head_hexsha": "0fa9ebeccc9177069aa09b2da84746b7532e3495", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "BatsResearch/taglets", "max_stars_repo_path": "taglets/scads/create/install.py", "max_stars_repo_stars_event_max_datetime": "2022-03-30T22:56:52.000Z", "max_stars_repo_stars_event_min_datetime": "2021-11-10T13:17:10.000Z", "num_tokens": 4389, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 19843 }
import csv import numpy as np import cv2 def read_data(file_path,header=True,delimiter=','): # The read-in data should be a N*W matrix, # where N is the length of the time sequences, # W is the number of sensors/data features i = 0 with open(file_path, 'r') as file: reader = csv.reader(file, delimiter = delimiter) data=[] for line in reader: if i == 0 and header: i += +1 else: line = np.array(line, dtype = 'float') # str2float if i == 0 or (i == 1 and header): data = line else: data = np.vstack((data, line)) i += 1 return data def read_binary(file_path,header=True,delimiter=','): # The read-in data should be a N*W matrix, # where N is the length of the time sequences, # W is the number of sensors/data features i = 0 with open(file_path, 'r') as file: reader = csv.reader(file, delimiter = delimiter) data=[] for line in reader: if i == 0 and header: i += +1 else: for j, element in enumerate(line): if element == 'True': line[j] = True elif element == 'False': line[j] = False else: raise ValueError("Data type is not boolean!!") line = np.array(line) # str2float if i == 0 or (i == 1 and header): data = line else: data = np.vstack((data, line)) i += 1 return data def read_human_data(file_path,num_header=2,delimiter='\t'): data = [] with open(file_path, 'r') as file: reader = csv.reader(file, delimiter=delimiter) for i,line in enumerate(reader): if i >= num_header: act = int(line[1]) x = float(line[3]) y = float(line[5]) vx = float(line[7]) # vy = float(line[8]) ax = float(line[9]) # ay = float(line[11]) # yaw =float([line[13]]) # yaw_dot =float([line[14]]) if i == num_header: data = np.array([act,x,y,vx,ax]) else: data = np.vstack([data, np.array([act,x,y,vx,ax])]) return data
{ "alphanum_fraction": 0.4574383453, "author": null, "avg_line_length": 34.4383561644, "converted": null, "ext": "py", "file": null, "hexsha": "0081ce0f5a205f67c9645abccd2a6851af09777a", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2020-02-20T04:12:32.000Z", "max_forks_repo_forks_event_min_datetime": "2020-02-20T04:12:32.000Z", "max_forks_repo_head_hexsha": "b5a720d59ef6243e32e97c544755fca555e713bc", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "MoonBlvd/Mask_RCNN", "max_forks_repo_path": "data_reader.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "b5a720d59ef6243e32e97c544755fca555e713bc", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "MoonBlvd/Mask_RCNN", "max_issues_repo_path": "data_reader.py", "max_line_length": 71, "max_stars_count": null, "max_stars_repo_head_hexsha": "b5a720d59ef6243e32e97c544755fca555e713bc", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "MoonBlvd/Mask_RCNN", "max_stars_repo_path": "data_reader.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 579, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2514 }
\section{EDA Figures} \begin{figure}[!h] \begin{minipage}{.45\textwidth} \caption{Date of birth frequency.} \label{fig:birth_date_freq} \centering \scalebox{0.5}{\input{./img/birth_date_freq.pgf}} % \includegraphics[width=10]{./img/birth_date_freq.pgf} \end{minipage} %\end{figure} %\begin{figure}[!h] \begin{minipage}{.45\textwidth} \caption{Interest earned frequency (logarithmic scale).} \label{fig:interest_earned_freq} \centering % \input{./img/interest_earned_freq.pgf} \includegraphics[width=1.2\textwidth]{./img/interest_earned_freq.png} % \includegraphics{./img/interest_earned_freq.png} \end{minipage} %\end{figure} %\begin{figure}[!h] \begin{minipage}{.45\textwidth} \caption{Monthly work frequency.} \label{fig:monthly_work_freq} \centering \scalebox{0.5}{\input{./img/monthly_work_freq.pgf}} \end{minipage} \end{figure} \begin{figure}[!h] \caption{Sections of the cross-correlation matrix.} \label{fig:cross-matrix} % \begin{subfigure}[b]{\linewidth} \begin{minipage}{.5\textwidth} \centering % \caption{A subfigure}\label{fig:1a} \scalebox{0.22}{\input{./img/cross-matrix-0-0.pgf}} \end{minipage} % \end{subfigure} % \begin{subfigure}[b]{\linewidth} \begin{minipage}{.5\textwidth} \centering % \caption{A subfigure}\label{fig:1a} \scalebox{0.22}{\input{./img/cross-matrix-0-1.pgf}} \end{minipage} % \end{subfigure} \end{figure} \begin{figure}[!h] \ContinuedFloat % \begin{subfigure}[b]{\linewidth} \begin{minipage}{.5\textwidth} \centering % \caption{A subfigure}\label{fig:1a} \scalebox{0.22}{\input{./img/cross-matrix-0-1.pgf}} \end{minipage} % \end{subfigure} % \begin{subfigure}[b]{\linewidth} \begin{minipage}{.5\textwidth} \centering % \caption{A subfigure}\label{fig:1a} \scalebox{0.22}{\input{./img/cross-matrix-1-1.pgf}} \end{minipage} % \end{subfigure} \end{figure} \begin{figure}[!h] \ContinuedFloat % \begin{subfigure}[b]{\linewidth} \begin{minipage}{.5\textwidth} \centering % \caption{A subfigure}\label{fig:1a} \scalebox{0.22}{\input{./img/cross-matrix-1-2.pgf}} \end{minipage} % \end{subfigure} % \begin{subfigure}[b]{\linewidth} \begin{minipage}{.5\textwidth} \centering % \caption{A subfigure}\label{fig:1a} \scalebox{0.22}{\input{./img/cross-matrix-2-2.pgf}} \end{minipage} % \end{subfigure} \end{figure}
{ "alphanum_fraction": 0.634525661, "author": null, "avg_line_length": 29.2272727273, "converted": null, "ext": "tex", "file": null, "hexsha": "7578443ac5fa1d95d18d3905f5dfb926aa84a540", "include": null, "lang": "TeX", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "cc8ad76f139b3b2b92238b52f5cd2e8b7223e854", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "hcastilho/hack6", "max_forks_repo_path": "doc/report/tex/eda_figures.tex", "max_issues_count": null, "max_issues_repo_head_hexsha": "cc8ad76f139b3b2b92238b52f5cd2e8b7223e854", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "hcastilho/hack6", "max_issues_repo_path": "doc/report/tex/eda_figures.tex", "max_line_length": 73, "max_stars_count": null, "max_stars_repo_head_hexsha": "cc8ad76f139b3b2b92238b52f5cd2e8b7223e854", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "hcastilho/hack6", "max_stars_repo_path": "doc/report/tex/eda_figures.tex", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 908, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 2572 }
import random import numpy as np import geomstats.backend as gs import geomstats.tests from geomstats.geometry.poincare_half_space import PoincareHalfSpace from tests.data_generation import _OpenSetTestData, _RiemannianMetricTestData class PoincareHalfSpaceTestData(_OpenSetTestData): dim_list = random.sample(range(2, 5), 2) space_args_list = [(dim,) for dim in dim_list] shape_list = [(dim,) for dim in dim_list] n_points_list = random.sample(range(2, 5), 2) n_vecs_list = random.sample(range(2, 5), 2) def belongs_test_data(self): smoke_data = [ dict(dim=2, vec=[1.5, 2.3], expected=True), dict(dim=2, vec=[[1.5, 2.0], [2.5, -0.3]], expected=[True, False]), ] return self.generate_tests(smoke_data) def half_space_to_ball_coordinates_test_data(self): smoke_data = [ dict(dim=2, point=[0.0, 1.0], expected=gs.zeros(2)), dict( dim=2, point=[[0.0, 1.0], [0.0, 2.0]], expected=[[0.0, 0.0], [0.0, 1.0 / 3.0]], ), ] return self.generate_tests(smoke_data) def ball_half_plane_tangent_are_inverse_test_data(self): smoke_data = [ dict( dim=2, tangent_vec=gs.array([0.5, 1.0]), base_point=gs.array([1.5, 2.3]), ) ] return self.generate_tests(smoke_data) def ball_to_half_space_coordinates_test_data(self): smoke_data = [dict(dim=2, point_ball=gs.array([-0.3, 0.7]))] return self.generate_tests(smoke_data) def half_space_coordinates_ball_coordinates_composition_test_data(self): smoke_data = [dict(dim=2, point_half_space=gs.array([1.5, 2.3]))] return self.generate_tests(smoke_data) def random_point_belongs_test_data(self): smoke_space_args_list = [(2,), (3,)] smoke_n_points_list = [1, 2] return self._random_point_belongs_test_data( smoke_space_args_list, smoke_n_points_list, self.space_args_list, self.n_points_list, ) def projection_belongs_test_data(self): return self._projection_belongs_test_data( self.space_args_list, self.shape_list, self.n_points_list ) def to_tangent_is_tangent_test_data(self): return self._to_tangent_is_tangent_test_data( PoincareHalfSpace, self.space_args_list, self.shape_list, self.n_vecs_list, ) def random_tangent_vec_is_tangent_test_data(self): return self._random_tangent_vec_is_tangent_test_data( PoincareHalfSpace, self.space_args_list, self.n_vecs_list ) def to_tangent_is_tangent_in_ambient_space_test_data(self): return self._to_tangent_is_tangent_in_ambient_space_test_data( PoincareHalfSpace, self.space_args_list, self.shape_list ) class PoincareHalfSpaceMetricTestData(_RiemannianMetricTestData): dim_list = random.sample(range(2, 5), 2) metric_args_list = [(dim,) for dim in dim_list] shape_list = [(dim,) for dim in dim_list] space_list = [PoincareHalfSpace(dim) for dim in dim_list] n_points_list = random.sample(range(1, 5), 2) n_tangent_vecs_list = random.sample(range(1, 5), 2) n_points_a_list = random.sample(range(1, 5), 2) n_points_b_list = [1] alpha_list = [1] * 2 n_rungs_list = [1] * 2 scheme_list = ["pole"] * 2 def inner_product_test_data(self): smoke_data = [ dict( dim=2, tangent_vec_a=[[1.0, 2.0], [3.0, 4.0]], tangent_vec_b=[[1.0, 2.0], [3.0, 4.0]], base_point=[[0.0, 1.0], [0.0, 5.0]], expected=[5.0, 1.0], ) ] return self.generate_tests(smoke_data) def exp_and_coordinates_tangent_test_data(self): smoke_data = [ dict( dim=2, tangent_vec=gs.array([0.0, 1.0]), base_point=gs.array([1.5, 2.3]), ) ] return self.generate_tests(smoke_data) def exp_test_data(self): def _exp(tangent_vec, base_point): circle_center = ( base_point[0] + base_point[1] * tangent_vec[1] / tangent_vec[0] ) circle_radius = gs.sqrt( (circle_center - base_point[0]) ** 2 + base_point[1] ** 2 ) moebius_d = 1 moebius_c = 1 / (2 * circle_radius) moebius_b = circle_center - circle_radius moebius_a = (circle_center + circle_radius) * moebius_c point_complex = base_point[0] + 1j * base_point[1] tangent_vec_complex = tangent_vec[0] + 1j * tangent_vec[1] point_moebius = ( 1j * (moebius_d * point_complex - moebius_b) / (moebius_c * point_complex - moebius_a) ) tangent_vec_moebius = ( -1j * tangent_vec_complex * (1j * moebius_c * point_moebius + moebius_d) ** 2 ) end_point_moebius = point_moebius * gs.exp( tangent_vec_moebius / point_moebius ) end_point_complex = (moebius_a * 1j * end_point_moebius + moebius_b) / ( moebius_c * 1j * end_point_moebius + moebius_d ) end_point_expected = gs.hstack( [np.real(end_point_complex), np.imag(end_point_complex)] ) return end_point_expected inputs_to_exp = [(gs.array([2.0, 1.0]), gs.array([1.0, 1.0]))] smoke_data = [] if not geomstats.tests.tf_backend(): for tangent_vec, base_point in inputs_to_exp: smoke_data.append( dict( dim=2, tangent_vec=tangent_vec, base_point=base_point, expected=_exp(tangent_vec, base_point), ) ) return self.generate_tests(smoke_data) def exp_shape_test_data(self): return self._exp_shape_test_data( self.metric_args_list, self.space_list, self.shape_list ) def log_shape_test_data(self): return self._log_shape_test_data( self.metric_args_list, self.space_list, ) def squared_dist_is_symmetric_test_data(self): return self._squared_dist_is_symmetric_test_data( self.metric_args_list, self.space_list, self.n_points_a_list, self.n_points_b_list, atol=gs.atol * 1000, ) def exp_belongs_test_data(self): return self._exp_belongs_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, belongs_atol=gs.atol * 10000, ) def log_is_tangent_test_data(self): return self._log_is_tangent_test_data( self.metric_args_list, self.space_list, self.n_points_list, is_tangent_atol=gs.atol * 10900, ) def geodesic_ivp_belongs_test_data(self): return self._geodesic_ivp_belongs_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_points_list, belongs_atol=gs.atol * 1000, ) def geodesic_bvp_belongs_test_data(self): return self._geodesic_bvp_belongs_test_data( self.metric_args_list, self.space_list, self.n_points_list, belongs_atol=gs.atol * 1000, ) def exp_after_log_test_data(self): return self._exp_after_log_test_data( self.metric_args_list, self.space_list, self.n_points_list, rtol=gs.rtol * 100, atol=gs.atol * 10000, ) def log_after_exp_test_data(self): return self._log_after_exp_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, rtol=gs.rtol * 100, atol=gs.atol * 10000, ) def exp_ladder_parallel_transport_test_data(self): return self._exp_ladder_parallel_transport_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, self.n_rungs_list, self.alpha_list, self.scheme_list, ) def exp_geodesic_ivp_test_data(self): return self._exp_geodesic_ivp_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, self.n_points_list, rtol=gs.rtol * 100000, atol=gs.atol * 100000, ) def parallel_transport_ivp_is_isometry_test_data(self): return self._parallel_transport_ivp_is_isometry_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, is_tangent_atol=gs.atol * 1000, atol=gs.atol * 1000, ) def parallel_transport_bvp_is_isometry_test_data(self): return self._parallel_transport_bvp_is_isometry_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_list, is_tangent_atol=gs.atol * 1000, atol=gs.atol * 1000, ) def dist_is_symmetric_test_data(self): return self._dist_is_symmetric_test_data( self.metric_args_list, self.space_list, self.n_points_a_list, self.n_points_b_list, ) def dist_is_positive_test_data(self): return self._dist_is_positive_test_data( self.metric_args_list, self.space_list, self.n_points_a_list, self.n_points_b_list, ) def squared_dist_is_positive_test_data(self): return self._squared_dist_is_positive_test_data( self.metric_args_list, self.space_list, self.n_points_a_list, self.n_points_b_list, ) def dist_is_norm_of_log_test_data(self): return self._dist_is_norm_of_log_test_data( self.metric_args_list, self.space_list, self.n_points_a_list, self.n_points_b_list, ) def dist_point_to_itself_is_zero_test_data(self): return self._dist_point_to_itself_is_zero_test_data( self.metric_args_list, self.space_list, self.n_points_list ) def triangle_inequality_of_dist_test_data(self): return self._triangle_inequality_of_dist_test_data( self.metric_args_list, self.space_list, self.n_points_list ) def inner_product_is_symmetric_test_data(self): return self._inner_product_is_symmetric_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, ) def retraction_lifting_test_data(self): return self._log_after_exp_test_data( self.metric_args_list, self.space_list, self.shape_list, self.n_tangent_vecs_list, rtol=gs.rtol * 100, atol=gs.atol * 10000, )
{ "alphanum_fraction": 0.5928700594, "author": null, "avg_line_length": 33.18, "converted": null, "ext": "py", "file": null, "hexsha": "540e863defc5ae7b6956a8d726f17b832b4c3c3f", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "414b6832c63df60788eda934c694d68fc939708f", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "chicken-biryani/geomstats", "max_forks_repo_path": "tests/data/poincare_half_space_data.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "414b6832c63df60788eda934c694d68fc939708f", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "chicken-biryani/geomstats", "max_issues_repo_path": "tests/data/poincare_half_space_data.py", "max_line_length": 84, "max_stars_count": null, "max_stars_repo_head_hexsha": "414b6832c63df60788eda934c694d68fc939708f", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "chicken-biryani/geomstats", "max_stars_repo_path": "tests/data/poincare_half_space_data.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2795, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 11613 }
[STATEMENT] lemma naturality_hom_induced: assumes "continuous_map X Y f" "f ` S \<subseteq> T" shows "hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S" [PROOF STATE] proof (prove) goal (1 subgoal): 1. hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S [PROOF STEP] proof (cases "q \<le> 0") [PROOF STATE] proof (state) goal (2 subgoals): 1. q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S 2. \<not> q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S [PROOF STEP] case False [PROOF STATE] proof (state) this: \<not> q \<le> 0 goal (2 subgoals): 1. q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S 2. \<not> q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S [PROOF STEP] then [PROOF STATE] proof (chain) picking this: \<not> q \<le> 0 [PROOF STEP] obtain p where p1: "p \<ge> Suc 0" and q: "q = int p" [PROOF STATE] proof (prove) using this: \<not> q \<le> 0 goal (1 subgoal): 1. (\<And>p. \<lbrakk>Suc 0 \<le> p; q = int p\<rbrakk> \<Longrightarrow> thesis) \<Longrightarrow> thesis [PROOF STEP] using zero_le_imp_eq_int [PROOF STATE] proof (prove) using this: \<not> q \<le> 0 0 \<le> ?k \<Longrightarrow> \<exists>n. ?k = int n goal (1 subgoal): 1. (\<And>p. \<lbrakk>Suc 0 \<le> p; q = int p\<rbrakk> \<Longrightarrow> thesis) \<Longrightarrow> thesis [PROOF STEP] by force [PROOF STATE] proof (state) this: Suc 0 \<le> p q = int p goal (2 subgoals): 1. q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S 2. \<not> q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S [PROOF STEP] show ?thesis [PROOF STATE] proof (prove) goal (1 subgoal): 1. hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S [PROOF STEP] proof [PROOF STATE] proof (state) goal (1 subgoal): 1. \<And>x. (hom_boundary q Y T \<circ> hom_induced q X S Y T f) x = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) x [PROOF STEP] fix c [PROOF STATE] proof (state) goal (1 subgoal): 1. \<And>x. (hom_boundary q Y T \<circ> hom_induced q X S Y T f) x = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) x [PROOF STEP] show "(hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c" [PROOF STATE] proof (prove) goal (1 subgoal): 1. (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] proof (cases "c \<in> carrier(relative_homology_group p X S)") [PROOF STATE] proof (state) goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] case True [PROOF STATE] proof (state) this: c \<in> carrier (relative_homology_group (int p) X S) goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] then [PROOF STATE] proof (chain) picking this: c \<in> carrier (relative_homology_group (int p) X S) [PROOF STEP] obtain a where ceq: "c = homologous_rel_set p X S a" and a: "singular_relcycle p X S a" [PROOF STATE] proof (prove) using this: c \<in> carrier (relative_homology_group (int p) X S) goal (1 subgoal): 1. (\<And>a. \<lbrakk>c = homologous_rel_set p X S a; singular_relcycle p X S a\<rbrakk> \<Longrightarrow> thesis) \<Longrightarrow> thesis [PROOF STEP] by (force simp: carrier_relative_homology_group) [PROOF STATE] proof (state) this: c = homologous_rel_set p X S a singular_relcycle p X S a goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] then [PROOF STATE] proof (chain) picking this: c = homologous_rel_set p X S a singular_relcycle p X S a [PROOF STEP] have sr: "singular_relcycle p Y T (chain_map p f a)" [PROOF STATE] proof (prove) using this: c = homologous_rel_set p X S a singular_relcycle p X S a goal (1 subgoal): 1. singular_relcycle p Y T (chain_map p f a) [PROOF STEP] using assms singular_relcycle_chain_map [PROOF STATE] proof (prove) using this: c = homologous_rel_set p X S a singular_relcycle p X S a continuous_map X Y f f ` S \<subseteq> T \<lbrakk>singular_relcycle ?p ?X ?S ?c; continuous_map ?X ?X' ?g; ?g ` ?S \<subseteq> ?T\<rbrakk> \<Longrightarrow> singular_relcycle ?p ?X' ?T (chain_map ?p ?g ?c) goal (1 subgoal): 1. singular_relcycle p Y T (chain_map p f a) [PROOF STEP] by fastforce [PROOF STATE] proof (state) this: singular_relcycle p Y T (chain_map p f a) goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] then [PROOF STATE] proof (chain) picking this: singular_relcycle p Y T (chain_map p f a) [PROOF STEP] have sb: "singular_relcycle (p - Suc 0) (subtopology X S) {} (chain_boundary p a)" [PROOF STATE] proof (prove) using this: singular_relcycle p Y T (chain_map p f a) goal (1 subgoal): 1. singular_relcycle (p - Suc 0) (subtopology X S) {} (chain_boundary p a) [PROOF STEP] by (metis One_nat_def a chain_boundary_boundary singular_chain_0 singular_relcycle) [PROOF STATE] proof (state) this: singular_relcycle (p - Suc 0) (subtopology X S) {} (chain_boundary p a) goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] have p1_eq: "int p - 1 = int (p - Suc 0)" [PROOF STATE] proof (prove) goal (1 subgoal): 1. int p - 1 = int (p - Suc 0) [PROOF STEP] using p1 [PROOF STATE] proof (prove) using this: Suc 0 \<le> p goal (1 subgoal): 1. int p - 1 = int (p - Suc 0) [PROOF STEP] by auto [PROOF STATE] proof (state) this: int p - 1 = int (p - Suc 0) goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] have cbm: "(chain_boundary p (chain_map p f a)) = (chain_map (p - Suc 0) f (chain_boundary p a))" [PROOF STATE] proof (prove) goal (1 subgoal): 1. chain_boundary p (chain_map p f a) = chain_map (p - Suc 0) f (chain_boundary p a) [PROOF STEP] using a chain_boundary_chain_map singular_relcycle [PROOF STATE] proof (prove) using this: singular_relcycle p X S a singular_chain ?p ?X ?c \<Longrightarrow> chain_boundary ?p (chain_map ?p ?g ?c) = chain_map (?p - Suc 0) ?g (chain_boundary ?p ?c) singular_relcycle ?p ?X ?S ?c = (singular_chain ?p ?X ?c \<and> singular_chain (?p - 1) (subtopology ?X ?S) (chain_boundary ?p ?c)) goal (1 subgoal): 1. chain_boundary p (chain_map p f a) = chain_map (p - Suc 0) f (chain_boundary p a) [PROOF STEP] by blast [PROOF STATE] proof (state) this: chain_boundary p (chain_map p f a) = chain_map (p - Suc 0) f (chain_boundary p a) goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] have contf: "continuous_map (subtopology X S) (subtopology Y T) f" [PROOF STATE] proof (prove) goal (1 subgoal): 1. continuous_map (subtopology X S) (subtopology Y T) f [PROOF STEP] using assms [PROOF STATE] proof (prove) using this: continuous_map X Y f f ` S \<subseteq> T goal (1 subgoal): 1. continuous_map (subtopology X S) (subtopology Y T) f [PROOF STEP] by (auto simp: continuous_map_in_subtopology topspace_subtopology continuous_map_from_subtopology) [PROOF STATE] proof (state) this: continuous_map (subtopology X S) (subtopology Y T) f goal (2 subgoals): 1. c \<in> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c 2. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] show ?thesis [PROOF STATE] proof (prove) goal (1 subgoal): 1. (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] unfolding q [PROOF STATE] proof (prove) goal (1 subgoal): 1. (hom_boundary (int p) Y T \<circ> hom_induced (int p) X S Y T f) c = (hom_induced (int p - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary (int p) X S) c [PROOF STEP] using assms p1 a [PROOF STATE] proof (prove) using this: continuous_map X Y f f ` S \<subseteq> T Suc 0 \<le> p singular_relcycle p X S a goal (1 subgoal): 1. (hom_boundary (int p) Y T \<circ> hom_induced (int p) X S Y T f) c = (hom_induced (int p - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary (int p) X S) c [PROOF STEP] apply (simp add: ceq assms hom_induced_chain_map hom_boundary_chain_boundary hom_boundary_chain_boundary [OF sr] singular_relcycle_def mod_subset_def) [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>Suc 0 \<le> p; singular_chain p X a \<and> singular_chain (p - Suc 0) (subtopology X S) (chain_boundary p a)\<rbrakk> \<Longrightarrow> homologous_rel_set (p - Suc 0) (subtopology Y T) {} (chain_boundary p (chain_map p f a)) = hom_induced (int p - 1) (subtopology X S) {} (subtopology Y T) {} f (homologous_rel_set (p - Suc 0) (subtopology X S) {} (chain_boundary p a)) [PROOF STEP] apply (simp add: p1_eq contf sb cbm hom_induced_chain_map) [PROOF STATE] proof (prove) goal: No subgoals! [PROOF STEP] done [PROOF STATE] proof (state) this: (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c goal (1 subgoal): 1. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] next [PROOF STATE] proof (state) goal (1 subgoal): 1. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] case False [PROOF STATE] proof (state) this: c \<notin> carrier (relative_homology_group (int p) X S) goal (1 subgoal): 1. c \<notin> carrier (relative_homology_group (int p) X S) \<Longrightarrow> (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] with assms [PROOF STATE] proof (chain) picking this: continuous_map X Y f f ` S \<subseteq> T c \<notin> carrier (relative_homology_group (int p) X S) [PROOF STEP] show ?thesis [PROOF STATE] proof (prove) using this: continuous_map X Y f f ` S \<subseteq> T c \<notin> carrier (relative_homology_group (int p) X S) goal (1 subgoal): 1. (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c [PROOF STEP] unfolding q o_def [PROOF STATE] proof (prove) using this: continuous_map X Y f f ` S \<subseteq> T c \<notin> carrier (relative_homology_group (int p) X S) goal (1 subgoal): 1. hom_boundary (int p) Y T (hom_induced (int p) X S Y T f c) = hom_induced (int p - 1) (subtopology X S) {} (subtopology Y T) {} f (hom_boundary (int p) X S c) [PROOF STEP] using assms [PROOF STATE] proof (prove) using this: continuous_map X Y f f ` S \<subseteq> T c \<notin> carrier (relative_homology_group (int p) X S) continuous_map X Y f f ` S \<subseteq> T goal (1 subgoal): 1. hom_boundary (int p) Y T (hom_induced (int p) X S Y T f c) = hom_induced (int p - 1) (subtopology X S) {} (subtopology Y T) {} f (hom_boundary (int p) X S c) [PROOF STEP] apply (simp add: hom_induced_default hom_boundary_default) [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>c \<notin> carrier (relative_homology_group (int p) X S); continuous_map X Y f; f ` S \<subseteq> T\<rbrakk> \<Longrightarrow> hom_boundary (int p) Y T (singular_relboundary_set p Y T) = hom_induced (int p - 1) (subtopology X S) {} (subtopology Y T) {} f \<one>\<^bsub>homology_group (int p - 1) (subtopology X S)\<^esub> [PROOF STEP] by (metis group_relative_homology_group hom_boundary hom_induced hom_one one_relative_homology_group) [PROOF STATE] proof (state) this: (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c goal: No subgoals! [PROOF STEP] qed [PROOF STATE] proof (state) this: (hom_boundary q Y T \<circ> hom_induced q X S Y T f) c = (hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S) c goal: No subgoals! [PROOF STEP] qed [PROOF STATE] proof (state) this: hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S goal (1 subgoal): 1. q \<le> 0 \<Longrightarrow> hom_boundary q Y T \<circ> hom_induced q X S Y T f = hom_induced (q - 1) (subtopology X S) {} (subtopology Y T) {} f \<circ> hom_boundary q X S [PROOF STEP] qed (force simp: hom_induced_trivial hom_boundary_trivial)
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": null, "hexsha": null, "include": null, "lang": null, "length": 48, "llama_tokens": 6985, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
import datetime import logging import sys from pytorch_lightning.core import datamodule from torch.utils import data #cd /home/dcast/adversarial_project ; /usr/bin/env /home/dcast/anaconda3/envs/deep_learning_torch/bin/python -- /home/dcast/adversarial_project/irt_to_nlp/train.py # sys.path.append("/content/adversarial_project") #to work in colab sys.path.append("/home/dcast/adversarial_project") import os import numpy as np import pytorch_lightning as pl import torch from pytorch_lightning.loggers import WandbLogger import wandb from openml.autotune import autotune_lr from irt_to_nlp.builders import build_dataset, get_callbacks, get_trainer,get_system from irt_to_nlp.config import CONFIG, create_config_dict import pandas as pd import uuid def apply_train_test(): def get_new_system_and_do_training_and_test(config,data_module,wandb_logger,callbacks,num_repeat=None,num_fold=None,run_test:bool=False): model=get_system(config,num_fold,num_repeat=num_repeat) # test_dataloader=data_module.test_dataloader() #create trainer callbacks[0].best_score=torch.tensor(np.Inf) trainer=get_trainer(wandb_logger,callbacks,config) result=trainer.fit(model,data_module) if run_test: result=trainer.test(model,test_dataloaders=data_module.test_dataloader()) return result return result # def generate_csv_results(results:list,data_name:str): # df=pd.DataFrame(results) # df.to_csv(f"{data_name}.csv") init_id=uuid.uuid1().int config=CONFIG() config_dict=create_config_dict(config) config_dict["id_group"]=init_id wandb.init( project='IRT-project-NLP', entity='dcastf01', config=config_dict) wandb_logger = WandbLogger( # offline=True, log_model=False ) config =wandb.config data_module=build_dataset(path_data_csv=config.path_data, dataset_name=config.dataset_name, batch_size=config.batch_size, model_name=config.model_name ) callbacks=get_callbacks(config,data_module) num_fold=config.num_fold if num_fold is None or num_fold==0: wandb.run.name=config.model_name[:5]+" "+\ datetime.datetime.utcnow().strftime("%Y-%m-%d %X") get_new_system_and_do_training_and_test(config,data_module,wandb_logger,callbacks,num_fold=num_fold,run_test=True) else: results=[] for num_repeat in range(config.repetitions): for fold in range(num_fold): print(f"Repeticion {num_repeat}, fold {fold}") if not num_repeat==0 or not fold==0: config=CONFIG() config_dict=create_config_dict(config) config_dict["id_group"]=init_id wandb.init( project='IRT-project-NLP', entity='dcastf01', config=config_dict) wandb_logger = WandbLogger( # offline=True, log_model=False ) config =wandb.config config=wandb.config wandb.run.name=str(num_repeat)+"_"+str(fold)+" "+config.model_name[:5]+" "+\ datetime.datetime.utcnow().strftime("%Y-%m-%d %X") # wandb.run.id_group=init_id result=get_new_system_and_do_training_and_test(config,data_module, wandb_logger,callbacks, num_repeat=num_repeat, num_fold=fold, run_test=False) if results: results.append(*result) wandb.finish() # config=CONFIG() # config_dict=create_config_dict(config) # wandb.init( # project='IRT-project-NLP', # entity='dcastf01', # config=config_dict) # wandb_logger = WandbLogger( # # offline=True, # log_model=False # ) # config =wandb.config # config=wandb.config # wandb.run.name="final"+config.model_name[:5]+" "+\ # datetime.datetime.utcnow().strftime("%Y-%m-%d %X") # # wandb.run.id_group=init_id # callbacks=get_callbacks(config,data_module,only_train_and_test=True) # result=get_new_system_and_do_training_and_test(config,data_module, # wandb_logger, # callbacks,num_fold=num_fold, # run_test=True # ) # results.append(*result) # generate_csv_results(results,config.dataset_name) # "construir lo del fold" def main(): os.environ["WANDB_IGNORE_GLOBS"]="*.ckpt" torch.manual_seed(0) print("empezando setup del experimento") torch.backends.cudnn.benchmark = True #aplicar todo lo del fold a partir de aquí apply_train_test() if __name__ == "__main__": main()
{ "alphanum_fraction": 0.539257981, "author": null, "avg_line_length": 37.6298701299, "converted": null, "ext": "py", "file": null, "hexsha": "52171e245698182223244a2290fe4a266c17a6f7", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "01564f7b4ff9f19021986e57f5bfad827213c8a6", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "dcastf01/creating_adversarial_images", "max_forks_repo_path": "irt_to_nlp/train.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "01564f7b4ff9f19021986e57f5bfad827213c8a6", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "dcastf01/creating_adversarial_images", "max_issues_repo_path": "irt_to_nlp/train.py", "max_line_length": 164, "max_stars_count": 1, "max_stars_repo_head_hexsha": "7b34cc3556a1b99ac67cb155fba8d0837c9b7b10", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "cmonserr/Why_Difficulty", "max_stars_repo_path": "irt_to_nlp/train.py", "max_stars_repo_stars_event_max_datetime": "2022-02-04T11:33:41.000Z", "max_stars_repo_stars_event_min_datetime": "2022-02-04T11:33:41.000Z", "num_tokens": 1121, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 5795 }
[STATEMENT] lemma OrdP_linear_lemma: assumes j: "atom j \<sharp> i" shows "{ OrdP (Var i) } \<turnstile> All j (OrdP (Var j) IMP (Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))" (is "_ \<turnstile> ?scheme") [PROOF STATE] proof (prove) goal (1 subgoal): 1. {OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] proof - [PROOF STATE] proof (state) goal (1 subgoal): 1. {OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] obtain k::name and l::name and m::name where k: "atom k \<sharp> (i,j)" and l: "atom l \<sharp> (i,j,k)" and m: "atom m \<sharp> (i,j)" [PROOF STATE] proof (prove) goal (1 subgoal): 1. (\<And>k l m. \<lbrakk>atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> thesis) \<Longrightarrow> thesis [PROOF STEP] by (metis obtain_fresh) [PROOF STATE] proof (state) this: atom k \<sharp> (i, j) atom l \<sharp> (i, j, k) atom m \<sharp> (i, j) goal (1 subgoal): 1. {OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] show ?thesis [PROOF STATE] proof (prove) goal (1 subgoal): 1. {OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] proof (rule OrdIndH [where i=i and j=k]) [PROOF STATE] proof (state) goal (2 subgoals): 1. atom k \<sharp> (i, SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) 2. {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) ((SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))(i::=Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] show "atom k \<sharp> (i, ?scheme)" [PROOF STATE] proof (prove) goal (1 subgoal): 1. atom k \<sharp> (i, SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] using k [PROOF STATE] proof (prove) using this: atom k \<sharp> (i, j) goal (1 subgoal): 1. atom k \<sharp> (i, SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] by (force simp add: fresh_Pair) [PROOF STATE] proof (state) this: atom k \<sharp> (i, SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) goal (1 subgoal): 1. {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) ((SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))(i::=Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] next [PROOF STATE] proof (state) goal (1 subgoal): 1. {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) ((SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))(i::=Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] show "{} \<turnstile> All i (OrdP (Var i) IMP (All2 k (Var i) (?scheme(i::= Var k)) IMP ?scheme))" [PROOF STATE] proof (prove) goal (1 subgoal): 1. {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) ((SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))(i::=Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] using j k [PROOF STATE] proof (prove) using this: atom j \<sharp> i atom k \<sharp> (i, j) goal (1 subgoal): 1. {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) ((SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))(i::=Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] apply simp [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) [PROOF STEP] apply (rule All_I Imp_I)+ [PROOF STATE] proof (prove) goal (3 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i IN Var j OR Var i EQ Var j OR Var j IN Var i 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>C\<in>{All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom j \<sharp> C 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>C\<in>{}. atom i \<sharp> C [PROOF STEP] defer 1 [PROOF STATE] proof (prove) goal (3 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>C\<in>{All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom j \<sharp> C 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>C\<in>{}. atom i \<sharp> C 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i IN Var j OR Var i EQ Var j OR Var j IN Var i [PROOF STEP] apply auto [2] [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i IN Var j OR Var i EQ Var j OR Var j IN Var i [PROOF STEP] apply (rule OrdIndH [where i=j and j=l]) [PROOF STATE] proof (prove) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom l \<sharp> (j, Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP All2 l (Var j) ((Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)(j::=Var l)) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] using l \<comment> \<open>nested induction\<close> [PROOF STATE] proof (prove) using this: atom l \<sharp> (i, j, k) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom l \<sharp> (j, Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP All2 l (Var j) ((Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)(j::=Var l)) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] apply (force simp add: fresh_Pair) [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP All2 l (Var j) ((Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)(j::=Var l)) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] apply simp [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) [PROOF STEP] apply (rule All_I Imp_I)+ [PROOF STATE] proof (prove) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i IN Var j OR Var i EQ Var j OR Var j IN Var i 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>C\<in>{All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom j \<sharp> C [PROOF STEP] prefer 2 [PROOF STATE] proof (prove) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>C\<in>{All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom j \<sharp> C 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i IN Var j OR Var i EQ Var j OR Var j IN Var i [PROOF STEP] apply force [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i IN Var j OR Var i EQ Var j OR Var j IN Var i [PROOF STEP] apply (rule Disj_3I) [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i EQ Var j [PROOF STEP] apply (rule Equality_I) \<comment> \<open>Now the opposite inclusion, @{term"Var j SUBS Var i"}\<close> [PROOF STATE] proof (prove) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var j SUBS Var i 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] apply (rule Subset_I [where i=m]) [PROOF STATE] proof (prove) goal (4 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var i 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom m \<sharp> (Var j, Var i) 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>B\<in>{Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom m \<sharp> B 4. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] apply (rule All2_E [THEN rotate4]) [PROOF STATE] proof (prove) goal (6 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom l \<sharp> Var j 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> ?x46 IN Var j 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {(Var i IN Var l OR Var i EQ Var l OR Var l IN Var i)(l::=?x46), Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var i 4. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom m \<sharp> (Var j, Var i) 5. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>B\<in>{Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom m \<sharp> B 6. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] using l m [PROOF STATE] proof (prove) using this: atom l \<sharp> (i, j, k) atom m \<sharp> (i, j) goal (6 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom l \<sharp> Var j 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> ?x46 IN Var j 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {(Var i IN Var l OR Var i EQ Var l OR Var l IN Var i)(l::=?x46), Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var i 4. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom m \<sharp> (Var j, Var i) 5. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>B\<in>{Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom m \<sharp> B 6. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] apply auto [PROOF STATE] proof (prove) goal (3 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var i IN Var m, Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var i 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var i EQ Var m, Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var i 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] apply (blast intro: ContraProve [THEN rotate3] OrdP_Trans) [PROOF STATE] proof (prove) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var i EQ Var m, Var m IN Var j, Neg (Var i IN Var j), Neg (Var j IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var i 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] apply (blast intro: ContraProve [THEN rotate3] Mem_cong [OF Hyp Refl, THEN Iff_MP2_same]) \<comment> \<open>Now the opposite inclusion, @{term"Var i SUBS Var j"}\<close> [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var i SUBS Var j [PROOF STEP] apply (rule Subset_I [where i=m]) [PROOF STATE] proof (prove) goal (3 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var m IN Var i, Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)} \<turnstile> Var m IN Var j 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> atom m \<sharp> (Var i, Var j) 3. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> \<forall>B\<in>{Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), All2 k (Var i) (SyntaxN.All j (OrdP (Var j) IMP Var k IN Var j OR Var k EQ Var j OR Var j IN Var k)), OrdP (Var i)}. atom m \<sharp> B [PROOF STEP] apply (rule All2_E [THEN rotate6], auto) [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {SyntaxN.All j (OrdP (Var j) IMP Var m IN Var j OR Var m EQ Var j OR Var j IN Var m), Var m IN Var i, Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), OrdP (Var i)} \<turnstile> Var m IN Var j [PROOF STEP] apply (rule All_E [where x = "Var j"], auto) [PROOF STATE] proof (prove) goal (2 subgoals): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var m EQ Var j, Var m IN Var i, Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), OrdP (Var i)} \<turnstile> Var m IN Var j 2. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var j IN Var m, Var m IN Var i, Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), OrdP (Var i)} \<turnstile> Var m IN Var j [PROOF STEP] apply (blast intro: ContraProve [THEN rotate4] Mem_cong [OF Hyp Refl, THEN Iff_MP_same]) [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<lbrakk>atom j \<sharp> i; atom k \<sharp> (i, j); atom l \<sharp> (i, j, k); atom m \<sharp> (i, j)\<rbrakk> \<Longrightarrow> {Var j IN Var m, Var m IN Var i, Neg (Var i IN Var j), Neg (Var j IN Var i), All2 l (Var j) (Var i IN Var l OR Var i EQ Var l OR Var l IN Var i), OrdP (Var j), OrdP (Var i)} \<turnstile> Var m IN Var j [PROOF STEP] apply (blast intro: ContraProve [THEN rotate4] OrdP_Trans) [PROOF STATE] proof (prove) goal: No subgoals! [PROOF STEP] done [PROOF STATE] proof (state) this: {} \<turnstile> SyntaxN.All i (OrdP (Var i) IMP All2 k (Var i) ((SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i))(i::=Var k)) IMP SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i)) goal: No subgoals! [PROOF STEP] qed [PROOF STATE] proof (state) this: {OrdP (Var i)} \<turnstile> SyntaxN.All j (OrdP (Var j) IMP Var i IN Var j OR Var i EQ Var j OR Var j IN Var i) goal: No subgoals! [PROOF STEP] qed
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": "Incompleteness_Predicates", "hexsha": null, "include": null, "lang": null, "length": 38, "llama_tokens": 9154, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
# Problem: https://projecteuler.net/problem=367 # To run the code: `./problem367.sh` """ * Define notations as follows: . let e(i) represent the element at position i of a permutation . a permutation p of n elements is represented as follows p = [e(1) e(2) e(3) ... e(n)] . a permutation p of n elements is called n-cyclic if: e(e(e(...e(1)))) = 1 ^^^^^^^^^^ `e` appears n times . let (e(1) e(2) ... e(n)) represent an n-cyclic permutation. . define a histogram as a function f: N -> N such that: f(i) = j means the element i appears j times in the data . define a `permutation cycle histogram` of permutation p of length n as a function h: N -> N such that h(i) = j means the (i-cyclic) sub-permutations appears j times in the permutation p where the domain of h is {0, 1, ..., n} * Let classify a permutation p using the "permutation cycle histogram". * Observation: Each permutation p is a collections of cyclic sub-permutations. For example, permutation | cyclic subpermutations | permutation cycle histogram [3 1 2 4 5] | (1 2 3) (4 5) | [0 0 1 1 0 0] -> one 2-cyclic subpermutation, one 3 cyclic subpermutation [1 2 3 4 5] | (1) (2) (3) (4) (5) | [0 5 0 0 0 0] -> five 1-cyclic subpermutation [2 3 4 5 1] | (1 2 3 4 5) | [0 0 0 0 0 1] -> one 5-cyclic subpermutation * Let hash_p(histogram) be a bijection function between all "permutation cycle histograms" and their corresponding hash. Let hash(permutation) be hash_p(histogram) where histogram is the "permutation cycle histogram" of the permutation. * Consider permutations of N elements. Let X = [0 1 ... N-1] * The code is as follows, - problem367.sh: compiles and runs the solver. - problem367.cpp: generates all permutations of X and counts the number of appearances of each possible hash of permutation cycle histograms. - problem367.py: it does 3 things * generates the transisition matrix: - for each possible permutation cycle histogram, it generates a representative permutation that maps to such the histogram - for all ways of picking 3 elements and shuffling them, it determines the probability of transitioning to other permutation cycle histograms * uses a linear solver to solve the absorbing Matrix chain where: - the sorted permutation is the absorbing state * calculates the expectation: - The linear solver above returns the expected number of steps to reach the absorbing state for all possible starting states. - We have the distribution of starting states from problem367.cpp, so we can use the linearity of expectation to calculate the answer to the problem: Expectation = sum{(probability of starting at state S) * (the expected number of steps to reach the absorbing state from S) | for all S in all possible starting states} """ from sage.all import * import itertools from collections import defaultdict import numpy as np def round_up_to_nearest_log_2(n): exp = 0 curr = 1 while curr < n: curr = curr * 2 exp = exp + 1 return exp def mask(n_bits): return 2**n_bits - 1 N = 11 M = 3 BITS_PER_ELEMENT = round_up_to_nearest_log_2(N) ELEMENT_MASK = mask(BITS_PER_ELEMENT) def dfs1(curr_node, first_node, seq, visited): if visited[curr_node]: return 0 visited[curr_node] = True return 1 + dfs1(seq[curr_node], first_node, seq, visited) def find_cycle_length_started_at(first_node, seq, visited): return dfs1(first_node, first_node, seq, visited) def find_cycle_length_freqs(seq): length_freq = [0] * (N+1) visited = [False] * N for first_node in range(N): if not visited[first_node]: cycle_length = find_cycle_length_started_at(first_node, seq, visited) length_freq[cycle_length] += 1 return length_freq def hash_freq(freq): h = 0 msb_index = 0 for length in range(N+1): element = freq[length] h = h | (element << msb_index) msb_index += BITS_PER_ELEMENT return h def hash_seq(seq): return hash_freq(find_cycle_length_freqs(seq)) def decode_hash(h): length_freqs = [] for i in range(N+1): length_freqs.append(h & mask(ELEMENT_MASK)) h >>= BITS_PER_ELEMENT return length_freqs def dfs2(curr_node, depth, curr_sum, target_sum, path, ans): if curr_sum > target_sum: return elif curr_sum == target_sum: freq = [0] * (N+1) for length in path[:depth]: freq[length] += 1 ans.append(freq) return next_node_lowerbound = curr_node for i in range(max(curr_node, 1), target_sum - curr_sum+1): path[depth] = i dfs2(i, depth+1, curr_sum + i, target_sum, path, ans) def get_permutation(seq, permutation_index): p = [0] * len(seq) for p_i, seq_i in enumerate(permutation_index): p[p_i] = seq[seq_i] return p def left_rotate_by_k(seq, k): return seq[k:] + seq[:k] def construct_representative_seq(cycle_length_freq): seq = [] cycle_lengths = [] for length, freq in enumerate(cycle_length_freq): for i in range(freq): cycle_lengths.append(length) first_index = 0 for length in cycle_lengths: seq += get_permutation(list(range(first_index, first_index + length)), left_rotate_by_k(list(range(length)), k=1)) first_index += length return seq def find_next_states(curr_length_freq): next_states = defaultdict(lambda: 0) seq = construct_representative_seq(cycle_length_freq) for picked_idx in itertools.combinations(list(range(N)), M): #for permutation in [[0,1,2], [0,2,1], [1,0,2], [1,2,0], [2,0,1], [2,1,0]]: for permutation in itertools.permutations(list(range(M))): new_seq = seq[:] for i in range(M): new_seq[picked_idx[i]] = seq[picked_idx[permutation[i]]] h = hash_seq(new_seq) next_states[h] += 1 return next_states if __name__ == "__main__": curr_path = [0] * N cycle_length_freqs_sum_to_N = [] dfs2(0, 0, 0, N, curr_path, cycle_length_freqs_sum_to_N) hash_to_index = {} max_index = 0 stopping_state_index = 0 for cycle_length_freq in cycle_length_freqs_sum_to_N: h = hash_freq(cycle_length_freq) hash_to_index[h] = max_index if cycle_length_freq[1] == N: stopping_state_index = max_index max_index += 1 n = len(cycle_length_freqs_sum_to_N) T = matrix(QQ, n, n) for cycle_length_freq in cycle_length_freqs_sum_to_N: curr_state_hash = hash_freq(cycle_length_freq) next_states = find_next_states(cycle_length_freq) n_next_states = sum(next_states.values()) if hash_to_index[curr_state_hash] == stopping_state_index: #T[stopping_state_index, stopping_state_index] = 0 continue for next_state_hash, freq in next_states.items(): T[hash_to_index[curr_state_hash], hash_to_index[next_state_hash]] += QQ("{}/{}".format(freq, n_next_states)) S = matrix(QQ, 1, n) with open("p367_count.txt") as f: for line in f.readlines(): h, freq = list(map(int, line.strip().split(" "))) S[0, hash_to_index[h]] = QQ("{}/{}".format(freq, factorial(N))) # Absorbing Markov Chain ABSORBING_STATE = stopping_state_index rhs = matrix(QQ, n, 1, lambda i, j: 1) rhs[ABSORBING_STATE, 0] = 0 A = matrix(QQ, np.eye(n)) - T x = A.solve_right(rhs) expectation = sum(S * x)[0] print(ceil(expectation))
{ "alphanum_fraction": 0.621148632, "author": null, "avg_line_length": 37.2201834862, "converted": null, "ext": "py", "file": null, "hexsha": "283e73b0ce691c24c813b7de22fb52a1f8e34669", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-11-02T12:08:46.000Z", "max_forks_repo_forks_event_min_datetime": "2021-11-02T12:08:46.000Z", "max_forks_repo_head_hexsha": "6f434be429bd26f5d0f84f5ab0f5fa2bd677c790", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "takekoputa/project-euler", "max_forks_repo_path": "4th_100/problem367.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "6f434be429bd26f5d0f84f5ab0f5fa2bd677c790", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "takekoputa/project-euler", "max_issues_repo_path": "4th_100/problem367.py", "max_line_length": 189, "max_stars_count": null, "max_stars_repo_head_hexsha": "6f434be429bd26f5d0f84f5ab0f5fa2bd677c790", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "takekoputa/project-euler", "max_stars_repo_path": "4th_100/problem367.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2060, "path": null, "reason": "import numpy,from sage", "repo": null, "save_path": null, "sha": null, "size": 8114 }
/- Copyright (c) 2018 Johannes Hölzl. All rights reserved. Released under Apache 2.0 license as described in the file LICENSE. Authors: Johannes Hölzl, Jens Wagemaker -/ import algebra.big_operators.basic import algebra.divisibility import algebra.invertible /-! # Associated, prime, and irreducible elements. -/ variables {α : Type*} {β : Type*} {γ : Type*} {δ : Type*} theorem is_unit_iff_dvd_one [comm_monoid α] {x : α} : is_unit x ↔ x ∣ 1 := ⟨by rintro ⟨u, rfl⟩; exact ⟨_, u.mul_inv.symm⟩, λ ⟨y, h⟩, ⟨⟨x, y, h.symm, by rw [h, mul_comm]⟩, rfl⟩⟩ theorem is_unit_iff_forall_dvd [comm_monoid α] {x : α} : is_unit x ↔ ∀ y, x ∣ y := is_unit_iff_dvd_one.trans ⟨λ h y, h.trans (one_dvd _), λ h, h _⟩ theorem is_unit_of_dvd_unit {α} [comm_monoid α] {x y : α} (xy : x ∣ y) (hu : is_unit y) : is_unit x := is_unit_iff_dvd_one.2 $ xy.trans $ is_unit_iff_dvd_one.1 hu lemma is_unit_of_dvd_one [comm_monoid α] : ∀a ∣ 1, is_unit (a:α) | a ⟨b, eq⟩ := ⟨units.mk_of_mul_eq_one a b eq.symm, rfl⟩ lemma dvd_and_not_dvd_iff [comm_cancel_monoid_with_zero α] {x y : α} : x ∣ y ∧ ¬y ∣ x ↔ dvd_not_unit x y := ⟨λ ⟨⟨d, hd⟩, hyx⟩, ⟨λ hx0, by simpa [hx0] using hyx, ⟨d, mt is_unit_iff_dvd_one.1 (λ ⟨e, he⟩, hyx ⟨e, by rw [hd, mul_assoc, ← he, mul_one]⟩), hd⟩⟩, λ ⟨hx0, d, hdu, hdx⟩, ⟨⟨d, hdx⟩, λ ⟨e, he⟩, hdu (is_unit_of_dvd_one _ ⟨e, mul_left_cancel₀ hx0 $ by conv {to_lhs, rw [he, hdx]};simp [mul_assoc]⟩)⟩⟩ lemma pow_dvd_pow_iff [comm_cancel_monoid_with_zero α] {x : α} {n m : ℕ} (h0 : x ≠ 0) (h1 : ¬ is_unit x) : x ^ n ∣ x ^ m ↔ n ≤ m := begin split, { intro h, rw [← not_lt], intro hmn, apply h1, have : x ^ m * x ∣ x ^ m * 1, { rw [← pow_succ', mul_one], exact (pow_dvd_pow _ (nat.succ_le_of_lt hmn)).trans h }, rwa [mul_dvd_mul_iff_left, ← is_unit_iff_dvd_one] at this, apply pow_ne_zero m h0 }, { apply pow_dvd_pow } end section prime variables [comm_monoid_with_zero α] /-- prime element of a `comm_monoid_with_zero` -/ def prime (p : α) : Prop := p ≠ 0 ∧ ¬ is_unit p ∧ (∀a b, p ∣ a * b → p ∣ a ∨ p ∣ b) namespace prime variables {p : α} (hp : prime p) include hp lemma ne_zero : p ≠ 0 := hp.1 lemma not_unit : ¬ is_unit p := hp.2.1 lemma not_dvd_one : ¬ p ∣ 1 := mt (is_unit_of_dvd_one _) hp.not_unit lemma ne_one : p ≠ 1 := λ h, hp.2.1 (h.symm ▸ is_unit_one) lemma dvd_or_dvd (hp : prime p) {a b : α} (h : p ∣ a * b) : p ∣ a ∨ p ∣ b := hp.2.2 a b h lemma dvd_of_dvd_pow (hp : prime p) {a : α} {n : ℕ} (h : p ∣ a^n) : p ∣ a := begin induction n with n ih, { rw pow_zero at h, have := is_unit_of_dvd_one _ h, have := not_unit hp, contradiction }, rw pow_succ at h, cases dvd_or_dvd hp h with dvd_a dvd_pow, { assumption }, exact ih dvd_pow end lemma exists_mem_multiset_dvd {s : multiset α} : p ∣ s.prod → ∃ a ∈ s, p ∣ a := multiset.induction_on s (λ h, (hp.not_dvd_one h).elim) $ λ a s ih h, have p ∣ a * s.prod, by simpa using h, match hp.dvd_or_dvd this with | or.inl h := ⟨a, multiset.mem_cons_self a s, h⟩ | or.inr h := let ⟨a, has, h⟩ := ih h in ⟨a, multiset.mem_cons_of_mem has, h⟩ end lemma exists_mem_multiset_map_dvd {s : multiset β} {f : β → α} : p ∣ (s.map f).prod → ∃ a ∈ s, p ∣ f a := λ h, by simpa only [exists_prop, multiset.mem_map, exists_exists_and_eq_and] using hp.exists_mem_multiset_dvd h lemma exists_mem_finset_dvd {s : finset β} {f : β → α} : p ∣ s.prod f → ∃ i ∈ s, p ∣ f i := hp.exists_mem_multiset_map_dvd end prime @[simp] lemma not_prime_zero : ¬ prime (0 : α) := λ h, h.ne_zero rfl @[simp] lemma not_prime_one : ¬ prime (1 : α) := λ h, h.not_unit is_unit_one end prime lemma prime.left_dvd_or_dvd_right_of_dvd_mul [comm_cancel_monoid_with_zero α] {p : α} (hp : prime p) {a b : α} : a ∣ p * b → p ∣ a ∨ a ∣ b := begin rintro ⟨c, hc⟩, rcases hp.2.2 a c (hc ▸ dvd_mul_right _ _) with h | ⟨x, rfl⟩, { exact or.inl h }, { rw [mul_left_comm, mul_right_inj' hp.ne_zero] at hc, exact or.inr (hc.symm ▸ dvd_mul_right _ _) } end /-- `irreducible p` states that `p` is non-unit and only factors into units. We explicitly avoid stating that `p` is non-zero, this would require a semiring. Assuming only a monoid allows us to reuse irreducible for associated elements. -/ class irreducible [monoid α] (p : α) : Prop := (not_unit' : ¬ is_unit p) (is_unit_or_is_unit' : ∀a b, p = a * b → is_unit a ∨ is_unit b) namespace irreducible lemma not_unit [monoid α] {p : α} (hp : irreducible p) : ¬ is_unit p := hp.1 lemma is_unit_or_is_unit [monoid α] {p : α} (hp : irreducible p) {a b : α} (h : p = a * b) : is_unit a ∨ is_unit b := irreducible.is_unit_or_is_unit' a b h end irreducible lemma irreducible_iff [monoid α] {p : α} : irreducible p ↔ ¬ is_unit p ∧ ∀a b, p = a * b → is_unit a ∨ is_unit b := ⟨λ h, ⟨h.1, h.2⟩, λ h, ⟨h.1, h.2⟩⟩ @[simp] theorem not_irreducible_one [monoid α] : ¬ irreducible (1 : α) := by simp [irreducible_iff] @[simp] theorem not_irreducible_zero [monoid_with_zero α] : ¬ irreducible (0 : α) | ⟨hn0, h⟩ := have is_unit (0:α) ∨ is_unit (0:α), from h 0 0 ((mul_zero 0).symm), this.elim hn0 hn0 theorem irreducible.ne_zero [monoid_with_zero α] : ∀ {p:α}, irreducible p → p ≠ 0 | _ hp rfl := not_irreducible_zero hp theorem of_irreducible_mul {α} [monoid α] {x y : α} : irreducible (x * y) → is_unit x ∨ is_unit y | ⟨_, h⟩ := h _ _ rfl theorem irreducible_or_factor {α} [monoid α] (x : α) (h : ¬ is_unit x) : irreducible x ∨ ∃ a b, ¬ is_unit a ∧ ¬ is_unit b ∧ a * b = x := begin haveI := classical.dec, refine or_iff_not_imp_right.2 (λ H, _), simp [h, irreducible_iff] at H ⊢, refine λ a b h, classical.by_contradiction $ λ o, _, simp [not_or_distrib] at o, exact H _ o.1 _ o.2 h.symm end protected lemma prime.irreducible [comm_cancel_monoid_with_zero α] {p : α} (hp : prime p) : irreducible p := ⟨hp.not_unit, λ a b hab, (show a * b ∣ a ∨ a * b ∣ b, from hab ▸ hp.dvd_or_dvd (hab ▸ dvd_rfl)).elim (λ ⟨x, hx⟩, or.inr (is_unit_iff_dvd_one.2 ⟨x, mul_right_cancel₀ (show a ≠ 0, from λ h, by simp [*, prime] at *) $ by conv {to_lhs, rw hx}; simp [mul_comm, mul_assoc, mul_left_comm]⟩)) (λ ⟨x, hx⟩, or.inl (is_unit_iff_dvd_one.2 ⟨x, mul_right_cancel₀ (show b ≠ 0, from λ h, by simp [*, prime] at *) $ by conv {to_lhs, rw hx}; simp [mul_comm, mul_assoc, mul_left_comm]⟩))⟩ lemma succ_dvd_or_succ_dvd_of_succ_sum_dvd_mul [comm_cancel_monoid_with_zero α] {p : α} (hp : prime p) {a b : α} {k l : ℕ} : p ^ k ∣ a → p ^ l ∣ b → p ^ ((k + l) + 1) ∣ a * b → p ^ (k + 1) ∣ a ∨ p ^ (l + 1) ∣ b := λ ⟨x, hx⟩ ⟨y, hy⟩ ⟨z, hz⟩, have h : p ^ (k + l) * (x * y) = p ^ (k + l) * (p * z), by simpa [mul_comm, pow_add, hx, hy, mul_assoc, mul_left_comm] using hz, have hp0: p ^ (k + l) ≠ 0, from pow_ne_zero _ hp.ne_zero, have hpd : p ∣ x * y, from ⟨z, by rwa [mul_right_inj' hp0] at h⟩, (hp.dvd_or_dvd hpd).elim (λ ⟨d, hd⟩, or.inl ⟨d, by simp [*, pow_succ, mul_comm, mul_left_comm, mul_assoc]⟩) (λ ⟨d, hd⟩, or.inr ⟨d, by simp [*, pow_succ, mul_comm, mul_left_comm, mul_assoc]⟩) /-- If `p` and `q` are irreducible, then `p ∣ q` implies `q ∣ p`. -/ lemma irreducible.dvd_symm [monoid α] {p q : α} (hp : irreducible p) (hq : irreducible q) : p ∣ q → q ∣ p := begin tactic.unfreeze_local_instances, rintros ⟨q', rfl⟩, rw is_unit.mul_right_dvd (or.resolve_left (of_irreducible_mul hq) hp.not_unit), end lemma irreducible.dvd_comm [monoid α] {p q : α} (hp : irreducible p) (hq : irreducible q) : p ∣ q ↔ q ∣ p := ⟨hp.dvd_symm hq, hq.dvd_symm hp⟩ /-- Two elements of a `monoid` are `associated` if one of them is another one multiplied by a unit on the right. -/ def associated [monoid α] (x y : α) : Prop := ∃u:units α, x * u = y local infix ` ~ᵤ ` : 50 := associated namespace associated @[refl] protected theorem refl [monoid α] (x : α) : x ~ᵤ x := ⟨1, by simp⟩ @[symm] protected theorem symm [monoid α] : ∀{x y : α}, x ~ᵤ y → y ~ᵤ x | x _ ⟨u, rfl⟩ := ⟨u⁻¹, by rw [mul_assoc, units.mul_inv, mul_one]⟩ @[trans] protected theorem trans [monoid α] : ∀{x y z : α}, x ~ᵤ y → y ~ᵤ z → x ~ᵤ z | x _ _ ⟨u, rfl⟩ ⟨v, rfl⟩ := ⟨u * v, by rw [units.coe_mul, mul_assoc]⟩ /-- The setoid of the relation `x ~ᵤ y` iff there is a unit `u` such that `x * u = y` -/ protected def setoid (α : Type*) [monoid α] : setoid α := { r := associated, iseqv := ⟨associated.refl, λa b, associated.symm, λa b c, associated.trans⟩ } end associated local attribute [instance] associated.setoid theorem unit_associated_one [monoid α] {u : units α} : (u : α) ~ᵤ 1 := ⟨u⁻¹, units.mul_inv u⟩ theorem associated_one_iff_is_unit [monoid α] {a : α} : (a : α) ~ᵤ 1 ↔ is_unit a := iff.intro (assume h, let ⟨c, h⟩ := h.symm in h ▸ ⟨c, (one_mul _).symm⟩) (assume ⟨c, h⟩, associated.symm ⟨c, by simp [h]⟩) theorem associated_zero_iff_eq_zero [monoid_with_zero α] (a : α) : a ~ᵤ 0 ↔ a = 0 := iff.intro (assume h, let ⟨u, h⟩ := h.symm in by simpa using h.symm) (assume h, h ▸ associated.refl a) theorem associated_one_of_mul_eq_one [comm_monoid α] {a : α} (b : α) (hab : a * b = 1) : a ~ᵤ 1 := show (units.mk_of_mul_eq_one a b hab : α) ~ᵤ 1, from unit_associated_one theorem associated_one_of_associated_mul_one [comm_monoid α] {a b : α} : a * b ~ᵤ 1 → a ~ᵤ 1 | ⟨u, h⟩ := associated_one_of_mul_eq_one (b * u) $ by simpa [mul_assoc] using h lemma associated_mul_unit_left {β : Type*} [monoid β] (a u : β) (hu : is_unit u) : associated (a * u) a := let ⟨u', hu⟩ := hu in ⟨u'⁻¹, hu ▸ units.mul_inv_cancel_right _ _⟩ lemma associated_unit_mul_left {β : Type*} [comm_monoid β] (a u : β) (hu : is_unit u) : associated (u * a) a := begin rw mul_comm, exact associated_mul_unit_left _ _ hu end lemma associated_mul_unit_right {β : Type*} [monoid β] (a u : β) (hu : is_unit u) : associated a (a * u) := (associated_mul_unit_left a u hu).symm lemma associated_unit_mul_right {β : Type*} [comm_monoid β] (a u : β) (hu : is_unit u) : associated a (u * a) := (associated_unit_mul_left a u hu).symm lemma associated.mul_mul [comm_monoid α] {a₁ a₂ b₁ b₂ : α} : a₁ ~ᵤ b₁ → a₂ ~ᵤ b₂ → (a₁ * a₂) ~ᵤ (b₁ * b₂) | ⟨c₁, h₁⟩ ⟨c₂, h₂⟩ := ⟨c₁ * c₂, by simp [h₁.symm, h₂.symm, mul_assoc, mul_comm, mul_left_comm]⟩ lemma associated.mul_left [comm_monoid α] (a : α) {b c : α} (h : b ~ᵤ c) : (a * b) ~ᵤ (a * c) := (associated.refl a).mul_mul h lemma associated.mul_right [comm_monoid α] {a b : α} (h : a ~ᵤ b) (c : α) : (a * c) ~ᵤ (b * c) := h.mul_mul (associated.refl c) lemma associated.pow_pow [comm_monoid α] {a b : α} {n : ℕ} (h : a ~ᵤ b) : a ^ n ~ᵤ b ^ n := begin induction n with n ih, { simp [h] }, convert h.mul_mul ih; rw pow_succ end protected lemma associated.dvd [monoid α] {a b : α} : a ~ᵤ b → a ∣ b := λ ⟨u, hu⟩, ⟨u, hu.symm⟩ protected lemma associated.dvd_dvd [monoid α] {a b : α} (h : a ~ᵤ b) : a ∣ b ∧ b ∣ a := ⟨h.dvd, h.symm.dvd⟩ theorem associated_of_dvd_dvd [cancel_monoid_with_zero α] {a b : α} (hab : a ∣ b) (hba : b ∣ a) : a ~ᵤ b := begin rcases hab with ⟨c, rfl⟩, rcases hba with ⟨d, a_eq⟩, by_cases ha0 : a = 0, { simp [*] at * }, have hac0 : a * c ≠ 0, { intro con, rw [con, zero_mul] at a_eq, apply ha0 a_eq, }, have : a * (c * d) = a * 1 := by rw [← mul_assoc, ← a_eq, mul_one], have hcd : (c * d) = 1, from mul_left_cancel₀ ha0 this, have : a * c * (d * c) = a * c * 1 := by rw [← mul_assoc, ← a_eq, mul_one], have hdc : d * c = 1, from mul_left_cancel₀ hac0 this, exact ⟨⟨c, d, hcd, hdc⟩, rfl⟩ end theorem dvd_dvd_iff_associated [cancel_monoid_with_zero α] {a b : α} : a ∣ b ∧ b ∣ a ↔ a ~ᵤ b := ⟨λ ⟨h1, h2⟩, associated_of_dvd_dvd h1 h2, associated.dvd_dvd⟩ lemma exists_associated_mem_of_dvd_prod [comm_cancel_monoid_with_zero α] {p : α} (hp : prime p) {s : multiset α} : (∀ r ∈ s, prime r) → p ∣ s.prod → ∃ q ∈ s, p ~ᵤ q := multiset.induction_on s (by simp [mt is_unit_iff_dvd_one.2 hp.not_unit]) (λ a s ih hs hps, begin rw [multiset.prod_cons] at hps, cases hp.dvd_or_dvd hps with h h, { use [a, by simp], cases h with u hu, cases (((hs a (multiset.mem_cons.2 (or.inl rfl))).irreducible) .is_unit_or_is_unit hu).resolve_left hp.not_unit with v hv, exact ⟨v, by simp [hu, hv]⟩ }, { rcases ih (λ r hr, hs _ (multiset.mem_cons.2 (or.inr hr))) h with ⟨q, hq₁, hq₂⟩, exact ⟨q, multiset.mem_cons.2 (or.inr hq₁), hq₂⟩ } end) lemma associated.dvd_iff_dvd_left [monoid α] {a b c : α} (h : a ~ᵤ b) : a ∣ c ↔ b ∣ c := let ⟨u, hu⟩ := h in hu ▸ units.mul_right_dvd.symm lemma associated.dvd_iff_dvd_right [monoid α] {a b c : α} (h : b ~ᵤ c) : a ∣ b ↔ a ∣ c := let ⟨u, hu⟩ := h in hu ▸ units.dvd_mul_right.symm lemma associated.eq_zero_iff [monoid_with_zero α] {a b : α} (h : a ~ᵤ b) : a = 0 ↔ b = 0 := ⟨λ ha, let ⟨u, hu⟩ := h in by simp [hu.symm, ha], λ hb, let ⟨u, hu⟩ := h.symm in by simp [hu.symm, hb]⟩ lemma associated.ne_zero_iff [monoid_with_zero α] {a b : α} (h : a ~ᵤ b) : a ≠ 0 ↔ b ≠ 0 := not_congr h.eq_zero_iff protected lemma associated.prime [comm_monoid_with_zero α] {p q : α} (h : p ~ᵤ q) (hp : prime p) : prime q := ⟨h.ne_zero_iff.1 hp.ne_zero, let ⟨u, hu⟩ := h in ⟨λ ⟨v, hv⟩, hp.not_unit ⟨v * u⁻¹, by simp [hv, hu.symm]⟩, hu ▸ by { simp [units.mul_right_dvd], intros a b, exact hp.dvd_or_dvd }⟩⟩ lemma irreducible.associated_of_dvd [cancel_monoid_with_zero α] {p q : α} (p_irr : irreducible p) (q_irr : irreducible q) (dvd : p ∣ q) : associated p q := associated_of_dvd_dvd dvd (p_irr.dvd_symm q_irr dvd) lemma irreducible.dvd_irreducible_iff_associated [cancel_monoid_with_zero α] {p q : α} (pp : irreducible p) (qp : irreducible q) : p ∣ q ↔ associated p q := ⟨irreducible.associated_of_dvd pp qp, associated.dvd⟩ lemma prime.associated_of_dvd [comm_cancel_monoid_with_zero α] {p q : α} (p_prime : prime p) (q_prime : prime q) (dvd : p ∣ q) : associated p q := p_prime.irreducible.associated_of_dvd q_prime.irreducible dvd theorem prime.dvd_prime_iff_associated [comm_cancel_monoid_with_zero α] {p q : α} (pp : prime p) (qp : prime q) : p ∣ q ↔ associated p q := pp.irreducible.dvd_irreducible_iff_associated qp.irreducible lemma associated.prime_iff [comm_monoid_with_zero α] {p q : α} (h : p ~ᵤ q) : prime p ↔ prime q := ⟨h.prime, h.symm.prime⟩ protected lemma associated.is_unit [monoid α] {a b : α} (h : a ~ᵤ b) : is_unit a → is_unit b := let ⟨u, hu⟩ := h in λ ⟨v, hv⟩, ⟨v * u, by simp [hv, hu.symm]⟩ lemma associated.is_unit_iff [monoid α] {a b : α} (h : a ~ᵤ b) : is_unit a ↔ is_unit b := ⟨h.is_unit, h.symm.is_unit⟩ protected lemma associated.irreducible [monoid α] {p q : α} (h : p ~ᵤ q) (hp : irreducible p) : irreducible q := ⟨mt h.symm.is_unit hp.1, let ⟨u, hu⟩ := h in λ a b hab, have hpab : p = a * (b * (u⁻¹ : units α)), from calc p = (p * u) * (u ⁻¹ : units α) : by simp ... = _ : by rw hu; simp [hab, mul_assoc], (hp.is_unit_or_is_unit hpab).elim or.inl (λ ⟨v, hv⟩, or.inr ⟨v * u, by simp [hv]⟩)⟩ protected lemma associated.irreducible_iff [monoid α] {p q : α} (h : p ~ᵤ q) : irreducible p ↔ irreducible q := ⟨h.irreducible, h.symm.irreducible⟩ lemma associated.of_mul_left [comm_cancel_monoid_with_zero α] {a b c d : α} (h : a * b ~ᵤ c * d) (h₁ : a ~ᵤ c) (ha : a ≠ 0) : b ~ᵤ d := let ⟨u, hu⟩ := h in let ⟨v, hv⟩ := associated.symm h₁ in ⟨u * (v : units α), mul_left_cancel₀ ha begin rw [← hv, mul_assoc c (v : α) d, mul_left_comm c, ← hu], simp [hv.symm, mul_assoc, mul_comm, mul_left_comm] end⟩ lemma associated.of_mul_right [comm_cancel_monoid_with_zero α] {a b c d : α} : a * b ~ᵤ c * d → b ~ᵤ d → b ≠ 0 → a ~ᵤ c := by rw [mul_comm a, mul_comm c]; exact associated.of_mul_left section unique_units variables [monoid α] [unique (units α)] lemma units_eq_one (u : units α) : u = 1 := subsingleton.elim u 1 theorem associated_iff_eq {x y : α} : x ~ᵤ y ↔ x = y := begin split, { rintro ⟨c, rfl⟩, rw [units_eq_one c, units.coe_one, mul_one] }, { rintro rfl, refl }, end theorem associated_eq_eq : (associated : α → α → Prop) = eq := by { ext, rw associated_iff_eq } end unique_units /-- The quotient of a monoid by the `associated` relation. Two elements `x` and `y` are associated iff there is a unit `u` such that `x * u = y`. There is a natural monoid structure on `associates α`. -/ def associates (α : Type*) [monoid α] : Type* := quotient (associated.setoid α) namespace associates open associated /-- The canonical quotient map from a monoid `α` into the `associates` of `α` -/ protected def mk {α : Type*} [monoid α] (a : α) : associates α := ⟦ a ⟧ instance [monoid α] : inhabited (associates α) := ⟨⟦1⟧⟩ theorem mk_eq_mk_iff_associated [monoid α] {a b : α} : associates.mk a = associates.mk b ↔ a ~ᵤ b := iff.intro quotient.exact quot.sound theorem quotient_mk_eq_mk [monoid α] (a : α) : ⟦ a ⟧ = associates.mk a := rfl theorem quot_mk_eq_mk [monoid α] (a : α) : quot.mk setoid.r a = associates.mk a := rfl theorem forall_associated [monoid α] {p : associates α → Prop} : (∀a, p a) ↔ (∀a, p (associates.mk a)) := iff.intro (assume h a, h _) (assume h a, quotient.induction_on a h) theorem mk_surjective [monoid α] : function.surjective (@associates.mk α _) := forall_associated.2 (λ a, ⟨a, rfl⟩) instance [monoid α] : has_one (associates α) := ⟨⟦ 1 ⟧⟩ theorem one_eq_mk_one [monoid α] : (1 : associates α) = associates.mk 1 := rfl instance [monoid α] : has_bot (associates α) := ⟨1⟩ lemma exists_rep [monoid α] (a : associates α) : ∃ a0 : α, associates.mk a0 = a := quot.exists_rep a section comm_monoid variable [comm_monoid α] instance : has_mul (associates α) := ⟨λa' b', quotient.lift_on₂ a' b' (λa b, ⟦ a * b ⟧) $ assume a₁ a₂ b₁ b₂ ⟨c₁, h₁⟩ ⟨c₂, h₂⟩, quotient.sound $ ⟨c₁ * c₂, by simp [h₁.symm, h₂.symm, mul_assoc, mul_comm, mul_left_comm]⟩⟩ theorem mk_mul_mk {x y : α} : associates.mk x * associates.mk y = associates.mk (x * y) := rfl instance : comm_monoid (associates α) := { one := 1, mul := (*), mul_one := assume a', quotient.induction_on a' $ assume a, show ⟦a * 1⟧ = ⟦ a ⟧, by simp, one_mul := assume a', quotient.induction_on a' $ assume a, show ⟦1 * a⟧ = ⟦ a ⟧, by simp, mul_assoc := assume a' b' c', quotient.induction_on₃ a' b' c' $ assume a b c, show ⟦a * b * c⟧ = ⟦a * (b * c)⟧, by rw [mul_assoc], mul_comm := assume a' b', quotient.induction_on₂ a' b' $ assume a b, show ⟦a * b⟧ = ⟦b * a⟧, by rw [mul_comm] } instance : preorder (associates α) := { le := has_dvd.dvd, le_refl := dvd_refl, le_trans := λ a b c, dvd_trans} @[simp] lemma mk_one : associates.mk (1 : α) = 1 := rfl /-- `associates.mk` as a `monoid_hom`. -/ protected def mk_monoid_hom : α →* (associates α) := ⟨associates.mk, mk_one, λ x y, mk_mul_mk⟩ @[simp] lemma mk_monoid_hom_apply (a : α) : associates.mk_monoid_hom a = associates.mk a := rfl lemma associated_map_mk {f : associates α →* α} (hinv : function.right_inverse f associates.mk) (a : α) : a ~ᵤ f (associates.mk a) := associates.mk_eq_mk_iff_associated.1 (hinv (associates.mk a)).symm lemma mk_pow (a : α) (n : ℕ) : associates.mk (a ^ n) = (associates.mk a) ^ n := by induction n; simp [*, pow_succ, associates.mk_mul_mk.symm] lemma dvd_eq_le : ((∣) : associates α → associates α → Prop) = (≤) := rfl theorem prod_mk {p : multiset α} : (p.map associates.mk).prod = associates.mk p.prod := multiset.induction_on p (by simp; refl) $ assume a s ih, by simp [ih]; refl theorem rel_associated_iff_map_eq_map {p q : multiset α} : multiset.rel associated p q ↔ p.map associates.mk = q.map associates.mk := by { rw [← multiset.rel_eq, multiset.rel_map], simp only [mk_eq_mk_iff_associated] } theorem mul_eq_one_iff {x y : associates α} : x * y = 1 ↔ (x = 1 ∧ y = 1) := iff.intro (quotient.induction_on₂ x y $ assume a b h, have a * b ~ᵤ 1, from quotient.exact h, ⟨quotient.sound $ associated_one_of_associated_mul_one this, quotient.sound $ associated_one_of_associated_mul_one $ by rwa [mul_comm] at this⟩) (by simp {contextual := tt}) theorem prod_eq_one_iff {p : multiset (associates α)} : p.prod = 1 ↔ (∀a ∈ p, (a:associates α) = 1) := multiset.induction_on p (by simp) (by simp [mul_eq_one_iff, or_imp_distrib, forall_and_distrib] {contextual := tt}) theorem units_eq_one (u : units (associates α)) : u = 1 := units.ext (mul_eq_one_iff.1 u.val_inv).1 instance unique_units : unique (units (associates α)) := { default := 1, uniq := associates.units_eq_one } theorem coe_unit_eq_one (u : units (associates α)): (u : associates α) = 1 := by simp theorem is_unit_iff_eq_one (a : associates α) : is_unit a ↔ a = 1 := iff.intro (assume ⟨u, h⟩, h ▸ coe_unit_eq_one _) (assume h, h.symm ▸ is_unit_one) theorem is_unit_mk {a : α} : is_unit (associates.mk a) ↔ is_unit a := calc is_unit (associates.mk a) ↔ a ~ᵤ 1 : by rw [is_unit_iff_eq_one, one_eq_mk_one, mk_eq_mk_iff_associated] ... ↔ is_unit a : associated_one_iff_is_unit section order theorem mul_mono {a b c d : associates α} (h₁ : a ≤ b) (h₂ : c ≤ d) : a * c ≤ b * d := let ⟨x, hx⟩ := h₁, ⟨y, hy⟩ := h₂ in ⟨x * y, by simp [hx, hy, mul_comm, mul_assoc, mul_left_comm]⟩ theorem one_le {a : associates α} : 1 ≤ a := dvd.intro _ (one_mul a) theorem prod_le_prod {p q : multiset (associates α)} (h : p ≤ q) : p.prod ≤ q.prod := begin haveI := classical.dec_eq (associates α), haveI := classical.dec_eq α, suffices : p.prod ≤ (p + (q - p)).prod, { rwa [add_tsub_cancel_of_le h] at this }, suffices : p.prod * 1 ≤ p.prod * (q - p).prod, { simpa }, exact mul_mono (le_refl p.prod) one_le end theorem le_mul_right {a b : associates α} : a ≤ a * b := ⟨b, rfl⟩ theorem le_mul_left {a b : associates α} : a ≤ b * a := by rw [mul_comm]; exact le_mul_right instance : order_bot (associates α) := { bot := 1, bot_le := assume a, one_le } end order end comm_monoid instance [has_zero α] [monoid α] : has_zero (associates α) := ⟨⟦ 0 ⟧⟩ instance [has_zero α] [monoid α] : has_top (associates α) := ⟨0⟩ section comm_monoid_with_zero variables [comm_monoid_with_zero α] @[simp] theorem mk_eq_zero {a : α} : associates.mk a = 0 ↔ a = 0 := ⟨assume h, (associated_zero_iff_eq_zero a).1 $ quotient.exact h, assume h, h.symm ▸ rfl⟩ theorem mk_ne_zero {a : α} : associates.mk a ≠ 0 ↔ a ≠ 0 := not_congr mk_eq_zero instance : comm_monoid_with_zero (associates α) := { zero_mul := by { rintro ⟨a⟩, show associates.mk (0 * a) = associates.mk 0, rw [zero_mul] }, mul_zero := by { rintro ⟨a⟩, show associates.mk (a * 0) = associates.mk 0, rw [mul_zero] }, .. associates.comm_monoid, .. associates.has_zero } instance : order_top (associates α) := { top := 0, le_top := assume a, ⟨0, (mul_zero a).symm⟩ } instance : bounded_order (associates α) := { .. associates.order_top, .. associates.order_bot } instance [nontrivial α] : nontrivial (associates α) := ⟨⟨0, 1, assume h, have (0 : α) ~ᵤ 1, from quotient.exact h, have (0 : α) = 1, from ((associated_zero_iff_eq_zero 1).1 this.symm).symm, zero_ne_one this⟩⟩ lemma exists_non_zero_rep {a : associates α} : a ≠ 0 → ∃ a0 : α, a0 ≠ 0 ∧ associates.mk a0 = a := quotient.induction_on a (λ b nz, ⟨b, mt (congr_arg quotient.mk) nz, rfl⟩) theorem dvd_of_mk_le_mk {a b : α} : associates.mk a ≤ associates.mk b → a ∣ b | ⟨c', hc'⟩ := (quotient.induction_on c' $ assume c hc, let ⟨d, hd⟩ := (quotient.exact hc).symm in ⟨(↑d) * c, calc b = (a * c) * ↑d : hd.symm ... = a * (↑d * c) : by ac_refl⟩) hc' theorem mk_le_mk_of_dvd {a b : α} : a ∣ b → associates.mk a ≤ associates.mk b := assume ⟨c, hc⟩, ⟨associates.mk c, by simp [hc]; refl⟩ theorem mk_le_mk_iff_dvd_iff {a b : α} : associates.mk a ≤ associates.mk b ↔ a ∣ b := iff.intro dvd_of_mk_le_mk mk_le_mk_of_dvd theorem mk_dvd_mk {a b : α} : associates.mk a ∣ associates.mk b ↔ a ∣ b := iff.intro dvd_of_mk_le_mk mk_le_mk_of_dvd lemma prime.le_or_le {p : associates α} (hp : prime p) {a b : associates α} (h : p ≤ a * b) : p ≤ a ∨ p ≤ b := hp.2.2 a b h lemma exists_mem_multiset_le_of_prime {s : multiset (associates α)} {p : associates α} (hp : prime p) : p ≤ s.prod → ∃a∈s, p ≤ a := multiset.induction_on s (assume ⟨d, eq⟩, (hp.ne_one (mul_eq_one_iff.1 eq.symm).1).elim) $ assume a s ih h, have p ≤ a * s.prod, by simpa using h, match prime.le_or_le hp this with | or.inl h := ⟨a, multiset.mem_cons_self a s, h⟩ | or.inr h := let ⟨a, has, h⟩ := ih h in ⟨a, multiset.mem_cons_of_mem has, h⟩ end lemma prime_mk (p : α) : prime (associates.mk p) ↔ _root_.prime p := begin rw [prime, _root_.prime, forall_associated], transitivity, { apply and_congr, refl, apply and_congr, refl, apply forall_congr, assume a, exact forall_associated }, apply and_congr mk_ne_zero, apply and_congr, { rw [is_unit_mk], }, apply forall_congr, assume a, apply forall_congr, assume b, rw [mk_mul_mk, mk_dvd_mk, mk_dvd_mk, mk_dvd_mk], end theorem irreducible_mk (a : α) : irreducible (associates.mk a) ↔ irreducible a := begin simp only [irreducible_iff, is_unit_mk], apply and_congr iff.rfl, split, { rintro h x y rfl, simpa [is_unit_mk] using h (associates.mk x) (associates.mk y) rfl }, { intros h x y, refine quotient.induction_on₂ x y (assume x y a_eq, _), rcases quotient.exact a_eq.symm with ⟨u, a_eq⟩, rw mul_assoc at a_eq, show is_unit (associates.mk x) ∨ is_unit (associates.mk y), simpa [is_unit_mk] using h _ _ a_eq.symm } end theorem mk_dvd_not_unit_mk_iff {a b : α} : dvd_not_unit (associates.mk a) (associates.mk b) ↔ dvd_not_unit a b := begin rw [dvd_not_unit, dvd_not_unit, mk_ne_zero], apply and_congr_right, intro ane0, split, { contrapose!, rw forall_associated, intros h x hx hbax, rw [mk_mul_mk, mk_eq_mk_iff_associated] at hbax, cases hbax with u hu, apply h (x * ↑u⁻¹), { rw is_unit_mk at hx, rw associated.is_unit_iff, apply hx, use u, simp, }, simp [← mul_assoc, ← hu] }, { rintro ⟨x, ⟨hx, rfl⟩⟩, use associates.mk x, simp [is_unit_mk, mk_mul_mk, hx], } end theorem dvd_not_unit_of_lt {a b : associates α} (hlt : a < b) : dvd_not_unit a b := begin split, { rintro rfl, apply not_lt_of_le _ hlt, apply dvd_zero }, rcases hlt with ⟨⟨x, rfl⟩, ndvd⟩, refine ⟨x, _, rfl⟩, contrapose! ndvd, rcases ndvd with ⟨u, rfl⟩, simp, end end comm_monoid_with_zero section comm_cancel_monoid_with_zero variable [comm_cancel_monoid_with_zero α] instance : partial_order (associates α) := { le_antisymm := λ a' b', quotient.induction_on₂ a' b' (λ a b hab hba, quot.sound $ associated_of_dvd_dvd (dvd_of_mk_le_mk hab) (dvd_of_mk_le_mk hba)) .. associates.preorder } instance : no_zero_divisors (associates α) := ⟨λ x y, (quotient.induction_on₂ x y $ assume a b h, have a * b = 0, from (associated_zero_iff_eq_zero _).1 (quotient.exact h), have a = 0 ∨ b = 0, from mul_eq_zero.1 this, this.imp (assume h, h.symm ▸ rfl) (assume h, h.symm ▸ rfl))⟩ theorem irreducible_iff_prime_iff : (∀ a : α, irreducible a ↔ prime a) ↔ (∀ a : (associates α), irreducible a ↔ prime a) := begin rw forall_associated, split; intros h a; have ha := h a; rw irreducible_mk at *; rw prime_mk at *; exact ha, end lemma eq_of_mul_eq_mul_left : ∀(a b c : associates α), a ≠ 0 → a * b = a * c → b = c := begin rintros ⟨a⟩ ⟨b⟩ ⟨c⟩ ha h, rcases quotient.exact' h with ⟨u, hu⟩, have hu : a * (b * ↑u) = a * c, { rwa [← mul_assoc] }, exact quotient.sound' ⟨u, mul_left_cancel₀ (mk_ne_zero.1 ha) hu⟩ end lemma eq_of_mul_eq_mul_right : ∀(a b c : associates α), b ≠ 0 → a * b = c * b → a = c := λ a b c bne0, (mul_comm b a) ▸ (mul_comm b c) ▸ (eq_of_mul_eq_mul_left b a c bne0) lemma le_of_mul_le_mul_left (a b c : associates α) (ha : a ≠ 0) : a * b ≤ a * c → b ≤ c | ⟨d, hd⟩ := ⟨d, eq_of_mul_eq_mul_left a _ _ ha $ by rwa ← mul_assoc⟩ lemma one_or_eq_of_le_of_prime : ∀(p m : associates α), prime p → m ≤ p → (m = 1 ∨ m = p) | _ m ⟨hp0, hp1, h⟩ ⟨d, rfl⟩ := match h m d dvd_rfl with | or.inl h := classical.by_cases (assume : m = 0, by simp [this]) $ assume : m ≠ 0, have m * d ≤ m * 1, by simpa using h, have d ≤ 1, from associates.le_of_mul_le_mul_left m d 1 ‹m ≠ 0› this, have d = 1, from bot_unique this, by simp [this] | or.inr h := classical.by_cases (assume : d = 0, by simp [this] at hp0; contradiction) $ assume : d ≠ 0, have d * m ≤ d * 1, by simpa [mul_comm] using h, or.inl $ bot_unique $ associates.le_of_mul_le_mul_left d m 1 ‹d ≠ 0› this end instance : comm_cancel_monoid_with_zero (associates α) := { mul_left_cancel_of_ne_zero := eq_of_mul_eq_mul_left, mul_right_cancel_of_ne_zero := eq_of_mul_eq_mul_right, .. (infer_instance : comm_monoid_with_zero (associates α)) } theorem dvd_not_unit_iff_lt {a b : associates α} : dvd_not_unit a b ↔ a < b := dvd_and_not_dvd_iff.symm end comm_cancel_monoid_with_zero end associates namespace multiset lemma prod_ne_zero_of_prime [comm_cancel_monoid_with_zero α] [nontrivial α] (s : multiset α) (h : ∀ x ∈ s, prime x) : s.prod ≠ 0 := multiset.prod_ne_zero (λ h0, prime.ne_zero (h 0 h0) rfl) end multiset
{ "alphanum_fraction": null, "author": "jjaassoonn", "avg_line_length": null, "converted": null, "ext": null, "file": null, "hexsha": null, "include": null, "lang": null, "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": "github-repos/lean/jjaassoonn-projective_space/projective_space-11fe19fe9d7991a272e7a40be4b6ad9b0c10c7ce/src/algebra/associated.lean", "reason": null, "repo": "projective_space", "save_path": "github-repos/lean/jjaassoonn-projective_space", "sha": "11fe19fe9d7991a272e7a40be4b6ad9b0c10c7ce", "size": null }
println("Fitting linear, time-invariant model") n,λ = 150,2.48 M2 = RegularizedLTIModel(n,N1,λ) lti = estfun(M2,p1,Q1)
{ "alphanum_fraction": 0.7166666667, "author": null, "avg_line_length": 20, "converted": null, "ext": "jl", "file": null, "hexsha": "7afba9996c1ecfd503c9b4e359e971734b956aab", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "649709a3feb8caed4e630a21f7b3b7349a93e9f0", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "wkearn/TidalDischargeModels.jl", "max_forks_repo_path": "test/lti.jl", "max_issues_count": 14, "max_issues_repo_head_hexsha": "649709a3feb8caed4e630a21f7b3b7349a93e9f0", "max_issues_repo_issues_event_max_datetime": "2017-10-18T16:43:55.000Z", "max_issues_repo_issues_event_min_datetime": "2017-03-11T19:24:39.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "wkearn/TidalDischargeModels.jl", "max_issues_repo_path": "test/lti.jl", "max_line_length": 47, "max_stars_count": null, "max_stars_repo_head_hexsha": "649709a3feb8caed4e630a21f7b3b7349a93e9f0", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "wkearn/TidalDischargeModels.jl", "max_stars_repo_path": "test/lti.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 48, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 120 }
""" Utiltiy helper functions of Machine Learning """ import numpy as np def sigmoid(val: np.ndarray) -> np.ndarray: """Sigmoid function .. math:: f(z) = \\frac{1}{1 + e^{-z}} Args: val (ndarray): input value Returns: np.ndarray: sigmoid value """ return 1 / (1 + np.exp(-val)) def moving_window_matrix(arr: np.ndarray, window: int, lag: int = 1) -> np.ndarray: """Create Moving Window matrix for 1D data. More details on this function. https://machinelearningexploration.readthedocs.io/en/latest/MathExploration/MovingWindow.html Args: arr (np.ndarray): input 1D array. window (int): window/ number of columns. lag (int, optional): lag count for moving. Defaults to 1. Returns: np.ndarray: transformed matrix. Raises: AssertionError: input array shape should be 1D like (m,). AssertionError: length of array should be greater than window size and lag. Example: >>> a = np.random.rand(100) >>> print(moving_window_matrix(a, 20, 2)) """ assert len(np.shape(arr)) == 1, 'input array shape should be 1D like (m,).' size = arr.shape[0] assert size > window and size > lag, \ 'length of array should be greater than window size and lag.' frame_width = size - window + 1 new_frame_width = int(np.ceil(frame_width / lag)) new_frame = np.empty(shape=(window, new_frame_width)) for row in range(0, window): new_frame[row] = arr[row: row+frame_width][::lag] return new_frame.T if __name__ == "__main__": # a = np.random.rand(100) # print(moving_window_matrix(a, 20, 2)) # import matplotlib.pyplot as plt # x = np.array([1, 2, 3, 4, 5, 6, 7, 8]) # y = np.array([1, 2, 3, 3, 4, 5, 7, 10]) # s, r, t = trend(x, y) # plt.plot(x, y, 'o', label='original') # plt.plot(x, t, '.-', label='regression line') # plt.legend() # plt.show() # x = np.arange(-10, 10) # y = x**2 + x**3 # s, r, l = polynomial_regression(x, y, 3) # plt.plot(x, y, 'ko', label='original') # plt.plot(x, l, '.-', label='regression line') # plt.legend() # plt.show() pass
{ "alphanum_fraction": 0.5892696123, "author": null, "avg_line_length": 24.9213483146, "converted": null, "ext": "py", "file": null, "hexsha": "699de411495f604474cfae0460531563b0b026bf", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "8219ae5cfc462e02f04bec6bdd7e3751b57d2a25", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "NishantBaheti/mightypy", "max_forks_repo_path": "src/mightypy/ml/_utils.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "8219ae5cfc462e02f04bec6bdd7e3751b57d2a25", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "NishantBaheti/mightypy", "max_issues_repo_path": "src/mightypy/ml/_utils.py", "max_line_length": 97, "max_stars_count": 1, "max_stars_repo_head_hexsha": "8219ae5cfc462e02f04bec6bdd7e3751b57d2a25", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "NishantBaheti/mightypy", "max_stars_repo_path": "src/mightypy/ml/_utils.py", "max_stars_repo_stars_event_max_datetime": "2022-02-03T19:32:45.000Z", "max_stars_repo_stars_event_min_datetime": "2022-02-03T19:32:45.000Z", "num_tokens": 644, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2218 }
struct LSQOutput β::Vector{Float64} ζ::Vector{Float64} Σβ::Matrix{Float64} Σζ::Matrix{Float64} Σβζ::Matrix{Float64} χ²::Float64 dof::Int βindex::Dict{Symbol, Union{Int, UnitRange{Int}}} zindex::Dict{Symbol, Union{Int, UnitRange{Int}}} end _names_ordered(index) = map(x->x[1], sort(collect(index), by=x->x[2])) """Return names of measured parameters `z`.""" names_z(output::LSQOutput) = _names_ordered(output.zindex) """Return names of unknown parameters `β`.""" names_β(output::LSQOutput) = _names_ordered(output.βindex) """Return standard uncertainty of unknown parameters `β`.""" uncertainty_β(output::LSQOutput) = [sqrt(output.Σβ[i, i]) for i in 1:length(output.β)] """Return covariance matrix of unknown parameters `β`.""" function covariance_β(output::LSQOutput) names = names_β(output) return NamedArray(output.Σβ, (names, names)) end """Return correlation matrix of unknown parameters `β`.""" function correlation_β(output::LSQOutput) cov = covariance_β(output) N = size(cov)[1] corr = similar(cov) for i in 1:N for j in 1:N corr[i, j] = cov[i, j]/√(cov[i, i]*cov[j, j]) end end return corr end """Display covariance matrix of unknown parameters `β`.""" function print_cov_β(output::LSQOutput) pretty_table( covariance_β(output); row_names=names_β(output), header=names_β(output), title="covariance of β parameters" ) end """Display correlation matrix of unknown parameters `β`.""" function print_corr_β(output::LSQOutput) pretty_table( correlation_β(output); row_names=names_β(output), header=names_β(output), title="correlation of β parameters" ) end
{ "alphanum_fraction": 0.6678141136, "author": null, "avg_line_length": 28.5737704918, "converted": null, "ext": "jl", "file": null, "hexsha": "3c6c4a0adabf3ce867c9ca4467a2902aac1d73f2", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "147f711ea957170ddb4e4d92a979c0887be9e252", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "nluetts/ConstrainedLeastSquares.jl", "max_forks_repo_path": "src/output.jl", "max_issues_count": null, "max_issues_repo_head_hexsha": "147f711ea957170ddb4e4d92a979c0887be9e252", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "nluetts/ConstrainedLeastSquares.jl", "max_issues_repo_path": "src/output.jl", "max_line_length": 86, "max_stars_count": null, "max_stars_repo_head_hexsha": "147f711ea957170ddb4e4d92a979c0887be9e252", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "nluetts/ConstrainedLeastSquares.jl", "max_stars_repo_path": "src/output.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 494, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 1743 }
""" This is an example of the application of DeepESN model for multivariate time-series prediction task on Piano-midi.de (see http://www-etud.iro.umontreal.ca/~boulanni/icml2012) dataset. The dataset is a polyphonic music task characterized by 88-dimensional sequences representing musical compositions. Starting from played notes at time t, the aim is to predict the played notes at time t+1. Reference paper for DeepESN model: C. Gallicchio, A. Micheli, L. Pedrelli, "Deep Reservoir Computing: A Critical Experimental Analysis", Neurocomputing, 2017, vol. 268, pp. 87-99 In this Example we consider the hyper-parameters designed in the following paper: C. Gallicchio, A. Micheli, L. Pedrelli, "Design of deep echo state networks", Neural Networks, 2018, vol. 108, pp. 33-47 """ from pathlib import Path import time import numpy as np from DeepESN import DeepESN from utils import computeMusicAccuracy, config_pianomidi, load_pianomidi, select_indexes class Struct(object): pass # sistemare indici per IP in config_pianomidi, mettere da un'altra parte # translation: set indexes by IP in config_plans, put elsewhere # this probably means the confusion of controlling how intrinsic plasticity is used # sistema selezione indici con transiente messi all'interno della rete # index selection system with transient placed inside the network # this probably means that the transient component in main is now redundant? def main(): # measure time for this code section t0 = time.perf_counter() # fix a seed for the reproducibility of results np.random.seed(7) # dataset path path = Path("data") (dataset, Nu, # dimension of a single data point # for example 88 for piano-midi.de # where 88 corresponds the number of keys on a piano error_function, optimization_problem, TR_indexes, # train set indices VL_indexes, # validation set indices TS_indexes # test set indices ) = load_pianomidi(path, computeMusicAccuracy) # load configuration for pianomidi task configs = config_pianomidi(list(TR_indexes) + list(VL_indexes)) # Be careful with memory usage # TODO: What does careful with memory usage mean? # What are the limits? Nr = 50 # number of recurrent units Nl = 1 # number of recurrent layers reg = 10.0**-2 # probably refers to lambda_r, readout regularization # BUG: however we also set regularization in the config file transient = 5 # initialize the ESN deepESN = DeepESN(Nu, Nr, Nl, configs, verbose=1) # states = deepESN.computeState(dataset.inputs, deepESN.IPconf.DeepIP, verbose=1) train_states = select_indexes(states, list(TR_indexes) + list(VL_indexes), transient) train_targets = select_indexes(dataset.targets, list(TR_indexes) + list(VL_indexes), transient) test_states = select_indexes(states, TS_indexes) test_targets = select_indexes(dataset.targets, TS_indexes) deepESN.trainReadout(train_states, train_targets, reg) train_outputs = deepESN.computeOutput(train_states) train_error = error_function(train_outputs, train_targets) print(f"Training ACC: {np.mean(train_error):0.5f} \n") test_outputs = deepESN.computeOutput(test_states) test_error = error_function(test_outputs, test_targets) print(f"Test ACC: {np.mean(test_error):0.5f} \n") # duration is difference between end time and start time t1 = time.perf_counter() print(f"Time elapsed: {t1-t0:0.5f} s") if __name__ == "__main__": main()
{ "alphanum_fraction": 0.7298127969, "author": null, "avg_line_length": 38.4838709677, "converted": null, "ext": "py", "file": null, "hexsha": "380269179ae781b7f8743a2681449f105fc5f807", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "40593b083c3b2ea9a547f4c925d66a50039f7340", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "Wehzie/master-thesis", "max_forks_repo_path": "src/main_MultivariatePolyphonic.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "40593b083c3b2ea9a547f4c925d66a50039f7340", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "Wehzie/master-thesis", "max_issues_repo_path": "src/main_MultivariatePolyphonic.py", "max_line_length": 115, "max_stars_count": null, "max_stars_repo_head_hexsha": "40593b083c3b2ea9a547f4c925d66a50039f7340", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "Wehzie/master-thesis", "max_stars_repo_path": "src/main_MultivariatePolyphonic.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 901, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 3579 }
''' #Requirement: opencv-python, tkinter # Usage instruction: Choose point (double click left mouse button) in the order leftUpper, RightUpper, LeftLower, RightLower Click "c" in the keyword for confirmation This is for two-chamber crylic transfer box, the dimension is two-chamber place preference _________445mm___________ | | | | 295mm | | | |________________________| fish tank _________500mm___________ | | | | 250mm | | | |________________________| water sink tank _________480mm___________ | | | | | | | | 480mm | | | | | | | |________________________| self stim _________250mm___________ | | | | 180mm | | | |________________________| Cross Maze Test 610 x 610 mm _ | | | | | | ___________| |___________ |__________ ___________| | | | | | | |_| water_searching_form_board _________640mm___________ | | | | 430mm | | | |________________________| homeCage _________280mm___________ | | | | 180mm | | | |________________________| The output video is 445 * 295 at 30fps, thus 1mm/pixel Changes can be made, for "out" and the "pts2" for other behavior test ''' import cv2 import numpy as np #import matplotlib.pyplot as plt import os from math import hypot from tkinter import Tk from tkinter.filedialog import askopenfilenames box_length = 500 box_width = 250 posList = [] def draw_circle(event,x,y,flags,param): global mouseX,mouseY if event == cv2.EVENT_LBUTTONDBLCLK: cv2.circle(img_raw,(x,y),1,(255,0,0),-1) mouseX,mouseY = x,y posList.append((x, y)) root1 = Tk() root1.withdraw() filez = askopenfilenames(parent = root1, title = 'Choose file') fourcc = cv2.VideoWriter_fourcc('m','p','4','v') file_order = 0; for fullFileName in root1.tk.splitlist(filez): print(fullFileName) filename = fullFileName (root, ext) =os.path.splitext(filename) duration = 1 # second freq = 440 # Hz file_order +=1 cap = cv2.VideoCapture(filename) i=0 while i<2: i+=1 ret, img_raw =cap.read() #start capture images from webcam if ret == False: break while i==1: cv2.namedWindow("image") cv2.setMouseCallback("image", draw_circle) #print(posList) posNp = np.array(posList) cv2.imshow('image',img_raw) k = cv2.waitKey(20) & 0xFF if k == ord("r"): ret, img_raw =cap.read() image = img_raw if k == ord("c"): break cap.release() cv2.destroyAllWindows() j=0 for fullFileName in root1.tk.splitlist(filez): j+=1 filename = fullFileName print(filename) (root, ext) =os.path.splitext(filename) cap = cv2.VideoCapture(filename) while not cap.isOpened(): cap = cv2.VideoCapture(filename) #os.system('play --no-show-progress --null --channels 1 synth %s sine %f' % (duration, freq)) print("Can't load the file") break fps = cap.get(cv2.CAP_PROP_FPS) #print(fps) duration = 1 # second freq = 440 # Hz cap = cv2.VideoCapture(filename) width = int(cap.get(3)) height = int(cap.get(4)) font = cv2.FONT_HERSHEY_SIMPLEX out = cv2.VideoWriter(root+'_GeoTran.mp4',fourcc,fps,(box_length,box_width)) while True: ret, img_raw =cap.read() #start capture images from webcam if ret == False: break img = img_raw rows,cols,ch = img.shape pts1 = np.float32(posList[(j-1)*4:(j*4)]) pts2 = np.float32([[0,0],[box_length,0],[0,box_width],[box_length,box_width]]) M = cv2.getPerspectiveTransform(pts1,pts2) dst = cv2.warpPerspective(img,M,(box_length,box_width)) out.write(dst) if cv2.waitKey(1) & 0xFF == ord('q'): break cap.release() out.release() cv2.destroyAllWindows()
{ "alphanum_fraction": 0.4995901639, "author": null, "avg_line_length": 26.2365591398, "converted": null, "ext": "py", "file": null, "hexsha": "55c392905745884931fe9fccacd3dea0f1f826f2", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "7e82025aff0691c7f36bbea38a97349468bc3c4e", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "li-shen-amy/behavior", "max_forks_repo_path": "mouse_tracking/python_code/Batch_GeoTran.py", "max_issues_count": 1, "max_issues_repo_head_hexsha": "7e82025aff0691c7f36bbea38a97349468bc3c4e", "max_issues_repo_issues_event_max_datetime": "2022-03-31T08:11:19.000Z", "max_issues_repo_issues_event_min_datetime": "2022-03-31T08:11:19.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "li-shen-amy/lick_detect", "max_issues_repo_path": "mouse_tracking/python_code/Batch_GeoTran.py", "max_line_length": 222, "max_stars_count": 1, "max_stars_repo_head_hexsha": "7e82025aff0691c7f36bbea38a97349468bc3c4e", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "li-shen-amy/lick_detect", "max_stars_repo_path": "mouse_tracking/python_code/Batch_GeoTran.py", "max_stars_repo_stars_event_max_datetime": "2021-02-26T00:02:24.000Z", "max_stars_repo_stars_event_min_datetime": "2021-02-26T00:02:24.000Z", "num_tokens": 1117, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4880 }
""" # WordSlots are the main objects of the grid, they are the word "spaces" to be filled """ import numpy as np class WordSlot(): def __init__(self, identifiant, length, first_letter_position, direction, initial_letters): self.identifiant = identifiant self.length = length self.first_letter_position = first_letter_position self.direction = direction self.slots = self.compute_slots() self.dic_crosses = {} self.current_letters = initial_letters self.number_of_possible_words = -1 self.current_possible_words = [] def set_current_possible_words(self, possible_words): """ Sets the list of words which can fit in the word slot given the letters already in place :param possible_words: [string] list of words which can fit in the word slot given the letters already in place :return: None """ self.current_possible_words = possible_words self.number_of_possible_words = len(possible_words) def what_current_word_would_be_with_an_added_letter(self,letter, position): """ Helper function to simulate the placement of a new letter :param letter: (char) letter we want to add to current word :param position: (int) position we want to add the letter to :return: (string) current word with the letter added """ new_word = self.current_letters.copy() new_word[position] = letter return new_word def check_if_word_is_filled(self): """ Helper function to check it current word is filled or if there are still some empty letters :return: None """ if "." not in self.current_letters: return True else: return False def propose_a_word_with_current_situation(self): """ Given the current letters, choose a possible word at random :return: (string) a possible word """ chosen_word = np.random.choice(self.current_possible_words) return chosen_word def set_a_word(self,word): """ Set a given word to be the word of this word slot :param word: (string) word to set :return: None """ self.current_letters = [x for x in word] self.number_of_possible_words = len(self.current_possible_words) @staticmethod def tab2string(tab): """ Helper function to convert an array to a string :param tab: array to convert :return: """ str_ = "" for item in tab: str_+= item return str_ def compute_crosses(self, temp_horizontal_slots_grid, temp_vertical_slots_grid): """ Compute crosses with other word slots (updates self.dic_crosses) :param temp_horizontal_slots_grid: ([int]) ID of horizontal words for each square :param temp_vertical_slots_grid: ([int]) ID of vertical words for each square :return: None """ relevant_grid_to_check = temp_horizontal_slots_grid if self.direction == "vertical" else \ temp_vertical_slots_grid for e,slot in enumerate(self.slots): corresponding_word_slot_id = relevant_grid_to_check[slot[0]][slot[1]] if corresponding_word_slot_id != "0": self.dic_crosses[int(corresponding_word_slot_id.split(".")[0])] = [e,int(corresponding_word_slot_id.split(".")[1])] def compute_slots(self): """ Compute position of spaces spanned by the word slot :return: ([int, int]) position of spaces spanned by the word slot """ slots = [] for i in range(self.length): if self.direction == "horizontal": slots.append([int(self.first_letter_position[0]), int(self.first_letter_position[1] + i)]) else: slots.append([int(self.first_letter_position[0] + i), int(self.first_letter_position[1])]) return slots
{ "alphanum_fraction": 0.6315142576, "author": null, "avg_line_length": 38.3773584906, "converted": null, "ext": "py", "file": null, "hexsha": "fa1ddf57466635b41ffb1203374fb3db16de4c85", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "a53907be76a65553ed3682f46e9a1b928ae48d75", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "clementdouarre/crosswords", "max_forks_repo_path": "src/WordSlot.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "a53907be76a65553ed3682f46e9a1b928ae48d75", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "clementdouarre/crosswords", "max_issues_repo_path": "src/WordSlot.py", "max_line_length": 132, "max_stars_count": null, "max_stars_repo_head_hexsha": "a53907be76a65553ed3682f46e9a1b928ae48d75", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "clementdouarre/crosswords", "max_stars_repo_path": "src/WordSlot.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 850, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4068 }
import argparse import json from pathlib import Path import numpy as np import torch import yaml from .binary_tree import BinaryTree from .graph import Graph from .random_walker import RandomWalker from .skipgram import SkipGram def run(args): if args.config_file.absolute().exists(): with open(args.config_file.absolute(), 'r') as config_io: hparams = yaml.load(config_io, Loader=yaml.FullLoader) else: raise FileNotFoundError(f"Config file not found. {args.config_file.absolute()}") graph = Graph( data_root=args.data_root, ) random_walker = RandomWalker( graph=graph, **hparams["random_walker"], ) binary_tree = BinaryTree( V=graph.V, n_dims=hparams["n_dims"] ) device = torch.device('cuda') if args.gpu else torch.device('cpu') skipgram = SkipGram( binary_tree=binary_tree, random_walker=random_walker, device=device, ** hparams["skipgram"], ) for epoch in range(hparams["random_walker"]["walks_per_node"]): if args.checkpoint_period > 0 and epoch % args.checkpoint_period == 0: checkpoint_root = args.output_root.joinpath('checkpoint', f'{epoch}') checkpoint_root.mkdir(parents=True, exist_ok=True) # save embeddings np.save( file=checkpoint_root.joinpath('Z.npy'), arr=binary_tree.get_node_embeddings().numpy(), ) # save context context = { 'optimizer': skipgram.optimizer.state_dict(), 'model': binary_tree.state_dict(), 'epoch': epoch, } torch.save(context, checkpoint_root.joinpath(f'{epoch}.pt')) skipgram.train() embeddings = binary_tree.get_node_embeddings() if not args.output_root.exists(): args.output_root.mkdir() np.save(Path(args.output_root).joinpath('Z.npy'), embeddings.cpu().numpy()) with open(args.output_root.joinpath('loss.json'), 'w') as io: json.dump(skipgram.loss_history, io) def get_parser(): parser = argparse.ArgumentParser( prog="deepwalk" ) parser.add_argument( "--data_root", type=Path, help="Path to the data root directory." ) parser.add_argument( "--config_file", type=Path, help="Path to the config file." ) parser.add_argument( "--output_root", type=Path, help="Path to the output root directory." ) parser.add_argument( "--checkpoint_period", type=int, default=0, help="Period of making checkpoint in epoch. 0 for no checkpoints." ) parser.add_argument('--gpu', action='store_true') return parser def main(): parser = get_parser() args = parser.parse_args() run(args) if __name__ == "__main__": main()
{ "alphanum_fraction": 0.6199236376, "author": null, "avg_line_length": 27.7019230769, "converted": null, "ext": "py", "file": null, "hexsha": "fd9e05b85996163e5cd68c3fdcdb8a89297ce51e", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "58325e000732e7588694a8ce97d1843e0192ba2c", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "helloybz/deepwalk_helloybz", "max_forks_repo_path": "deepwalk/__main__.py", "max_issues_count": 18, "max_issues_repo_head_hexsha": "58325e000732e7588694a8ce97d1843e0192ba2c", "max_issues_repo_issues_event_max_datetime": "2021-11-30T07:23:43.000Z", "max_issues_repo_issues_event_min_datetime": "2021-11-20T17:20:45.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "helloybz/deepwalk_helloybz", "max_issues_repo_path": "deepwalk/__main__.py", "max_line_length": 88, "max_stars_count": 1, "max_stars_repo_head_hexsha": "58325e000732e7588694a8ce97d1843e0192ba2c", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "helloybz/deepwalk_helloybz", "max_stars_repo_path": "deepwalk/__main__.py", "max_stars_repo_stars_event_max_datetime": "2021-11-23T12:34:32.000Z", "max_stars_repo_stars_event_min_datetime": "2021-11-23T12:34:32.000Z", "num_tokens": 631, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2881 }
import matplotlib matplotlib.use('pdf') from matplotlib import pyplot as plt from mpl_toolkits.axes_grid1 import make_axes_locatable from scipy.cluster.hierarchy import dendrogram, linkage from scipy.spatial.distance import is_valid_dm import numpy as np import general_scripts as gs matplotlib.rcParams['font.family'] = ['sans-serif'] matplotlib.rcParams['font.size'] = 10 #matplotlib.rcParams['mathtext.default'] = ['regular'] matplotlib.rcParams['text.usetex'] = True #params = {'mathtext.default': 'bf' } #plt.rcParams.update(params) matplotlib.rcParams['text.latex.preamble'] = [ r'\usepackage{siunitx}', # i need upright \micro symbols, but you need... r'\sisetup{detect-all}', # ...this to force siunitx to actually use your fonts r'\usepackage{helvet}', # set the normal font here r'\usepackage{sansmath}', # load up the sansmath so that math -> helvet r'\sansmath' # <- tricky! -- gotta actually tell tex to use! ] label_font0 = {'fontname':'Nimbus Mono L', 'fontsize':6, 'fontweight':'normal' } label_font1 = {'fontname':'Nimbus Mono L', 'fontsize':10, 'fontweight':'normal' } label_font2 = {'fontname':'Nimbus Mono L', 'fontsize':14, 'fontweight':'bold' } colorbar_min=0.0 colorbar_max=1.0 colorbar_max2=0.5 metric='V_R' # Adopted from gnuplot schemes #set palette defined (0 "black", 1 '#2233bb', 2 "yellow", 3 "red", 4 "pink", 5 "white") # i.e. #00-00-00 , #22-33-bb, #FF-FF-00, #FF-00-00, #FF-C0-CB, #FF-FF-FF cdict = {'red': [(0.0, 0.000, 0.000), (0.2, 0.133, 0.133), (0.4, 1.000, 1.000), (0.6, 0.800, 0.800), (0.8, 1.000, 1.000), (1.0, 1.000, 1.000)], 'green': [(0.0, 0.000, 0.000), (0.2, 0.199, 0.199), (0.4, 1.000, 1.000), (0.6, 0.000, 0.000), (0.8, 0.750, 0.750), (1.0, 1.000, 1.000)], 'blue': [(0.0, 0.000, 0.000), (0.2, 0.730, 0.730), (0.4, 0.000, 0.000), (0.6, 0.000, 0.000), (0.8, 0.793, 0.793), (1.0, 1.000, 1.000)] } #cdict = {'red': [(0.0, 0.0, 0.0), (1.0, 0.0, 0.0)], 'green': [(0.0, 1.0, 1.0), (1.0, 0.0, 0.0)], 'blue': [(0.0, 1.0, 1.0), (1.0, 0.0, 0.0)]} cmap_segmented = matplotlib.colors.LinearSegmentedColormap('Gnuplot', cdict) fig_dims=(5,5) dpi=300 #box_style = dict(boxstyle='round', facecolor='#FFEEDD') # place a text box in upper left in axes coords #ax.text(0.05, 0.95, textstr, transform=ax.transAxes, fontsize=14, # verticalalignment='top', bbox=props) rna_font = {'fontname':'Nimbus Mono L'} #plt.close('all') rna_labels=['A$_{10}$','U$_{10}$','U$_5$C$_5$','C$_5$U$_5$','C$_{10}$','U$_3$C$_4$U$_3$','C$_3$U$_4$C$_3$','U$_9$','U$_8$','U$_7$','U$_6$','U$_5$','UGU$_8$'] rna_labels_pos=range(len(rna_labels)) # Print matrix #figC, (ax1, ax2) = plt.subplots( 1, 2, figsize=(6, 3), dpi=300 ) figC = plt.figure( figsize=(5, 3), dpi=300 ) Xaverage=[] conc_list=['0.0','0.1','0.2','0.4','0.6','0.8','0.9','1.0'] #conc_list=['1.0'] numConc=len(conc_list) bFirst=True for e in range(numConc): in_file='./fitted_'+metric+'_'+conc_list[e]+'_matrix.dat' out_file='matrix_'+metric+'_'+conc_list[e]+'.pdf' X = gs.load_matrix(in_file) # Rescale and symmetrize matrix. Xp = np.maximum( np.zeros( X.shape ), X ) Xp = 0.5*( Xp + Xp.T ) print( is_valid_dm(Xp) ) if bFirst: bFirst=False Xaverage=Xp else: Xaverage=Xaverage+Xp plt.clf() ax1 = plt.subplot2grid((1,7), (0,0), colspan=5) ax2 = plt.subplot2grid((1,7), (0,5), colspan=2) plt.subplots_adjust(left=0.04, right=0.99, bottom=0.03, top=0.76, wspace=3.0, hspace=None) plt.figtext(s='Prot:RNA-ratio', x=0.68, y=0.95, horizontalalignment='center', verticalalignment='center', **label_font1 ) plt.figtext(s='1.0:%s' % conc_list[e], x=0.68, y=0.90, horizontalalignment='center', verticalalignment='center', **label_font2 ) #fig1 = plt.figure(figsize=(5,5), dpi=300 ) plt.sca(ax1) plt.cla() ax1.xaxis.tick_top() plt.xticks(rna_labels_pos, rna_labels, rotation='vertical', **rna_font) plt.yticks(rna_labels_pos, rna_labels, rotation='horizontal', **rna_font) graph = ax1.imshow(X,interpolation='none',cmap=cmap_segmented, vmin=colorbar_min, vmax=colorbar_max) divider = make_axes_locatable(ax1) cax = divider.append_axes("right", size="5%", pad=0.05) cbar = figC.colorbar(graph, cax=cax) cbar.ax.set_ylabel('pairwise reduced-$\chi$') #plt.margins(1.0, tight=False) plt.sca(ax2) plt.cla() ax2.xaxis.tick_top() ax2.xaxis.set_label_position('top') # plt.xlabel('cluster distance') plt.xlabel('$d_{ab}=0.5(\chi_{ab}+\chi_{ba})$', **label_font0) ax2.spines['left'].set_visible(False) ax2.spines['bottom'].set_visible(False) ax2.spines['right'].set_visible(False) # generate the linkage matrix for the dendrogram. Z = linkage(Xp, 'average') dendrogram(Z, labels=rna_labels, color_threshold=0.5*max(Z[:,2]), orientation='right', #leaf_rotation=90., # rotates the x axis labels leaf_font_size=10 ) figC.savefig(out_file) #print( inconsistent(Z) ) print( "= = Output for %s is complete." % conc_list[e] ) # Do final plot with aggregate data Xaverage /= numConc out_file='matrix_'+metric+'_aggregate.pdf' plt.clf() ax1 = plt.subplot2grid((1,7), (0,0), colspan=5) ax2 = plt.subplot2grid((1,7), (0,5), colspan=2) plt.subplots_adjust(left=0.04, right=0.99, bottom=0.03, top=0.76, wspace=3.0, hspace=None) plt.figtext(s='Aggregate', x=0.68, y=0.95, horizontalalignment='center', verticalalignment='center', **label_font1 ) #fig1 = plt.figure(figsize=(5,5), dpi=300 ) plt.sca(ax1) plt.cla() ax1.xaxis.tick_top() plt.xticks(rna_labels_pos, rna_labels, rotation='vertical', **rna_font) plt.yticks(rna_labels_pos, rna_labels, rotation='horizontal', **rna_font) graph = ax1.imshow(Xaverage,interpolation='none',cmap=cmap_segmented, vmin=colorbar_min, vmax=colorbar_max2) #plt.margins(1.0, tight=False) divider = make_axes_locatable(ax1) cax = divider.append_axes("right", size="5%", pad=0.05) cbar = figC.colorbar(graph, cax=cax) cbar.ax.set_ylabel('pairwise reduced-$\chi$') plt.sca(ax2) plt.cla() ax2.xaxis.tick_top() ax2.xaxis.set_label_position('top') plt.xlabel('cluster distance') plt.xlabel('$d_{ab}=0.5(\chi_{ab}+\chi_{ba})$', **label_font0) ax2.spines['left'].set_visible(False) ax2.spines['bottom'].set_visible(False) ax2.spines['right'].set_visible(False) # generate the linkage matrix for the dendrogram. Z = linkage(Xaverage, 'average') dendrogram(Z, labels=rna_labels, color_threshold=0.5*max(Z[:,2]), orientation='right', #leaf_rotation=90., # rotates the x axis labels leaf_font_size=10 ) figC.savefig(out_file)
{ "alphanum_fraction": 0.6265841014, "author": null, "avg_line_length": 37.3333333333, "converted": null, "ext": "py", "file": null, "hexsha": "31403b1302787ab14d366253681a621ce85ecd15", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "160666f51f024243b16774f2cfa3e130bcb5b1c8", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "zharmad/SAXScreen", "max_forks_repo_path": "scripts/plot-dendrogram-VR-example.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "160666f51f024243b16774f2cfa3e130bcb5b1c8", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "zharmad/SAXScreen", "max_issues_repo_path": "scripts/plot-dendrogram-VR-example.py", "max_line_length": 157, "max_stars_count": null, "max_stars_repo_head_hexsha": "160666f51f024243b16774f2cfa3e130bcb5b1c8", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "zharmad/SAXScreen", "max_stars_repo_path": "scripts/plot-dendrogram-VR-example.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2345, "path": null, "reason": "import numpy,from scipy", "repo": null, "save_path": null, "sha": null, "size": 6944 }
! ! -- LAPACK95 interface driver routine (version 3.0) -- ! UNI-C, Denmark; Univ. of Tennessee, USA; NAG Ltd., UK ! September, 2000 ! ! .. USE STATEMENTS .. USE LA_PRECISION, ONLY: WP => SP USE LA_AUXMOD, ONLY: ERINFO, LSAME ! .. IMPLICIT STATEMENT .. IMPLICIT NONE ! .. SCALAR ARGUMENTS .. CHARACTER(LEN=1), INTENT(IN), OPTIONAL :: JOBZ, UPLO INTEGER, INTENT(IN), OPTIONAL :: ITYPE INTEGER, INTENT(OUT), OPTIONAL :: INFO ! .. ARRAY ARGUMENTS .. COMPLEX(WP), INTENT(INOUT) :: A(:,:), B(:,:) REAL(WP), INTENT(OUT) :: W(:) !---------------------------------------------------------------------- ! ! Purpose ! ======= ! ! LA_SYGV, LA_SYGVD, LA_HEGV and LA_HEGVD compute all eigenvalues ! and, optionally, all eigenvectors of generalized eigenvalue problems of ! the form A*z = lambda*B*z, A*B*z = lambda*z, and B*A*z = lambda*z, ! where A and B are real symmetric in the cases of LA_SYGV and LA_SYGVD ! and complex Hermitian in the cases of LA_HEGV and LA_HEGVD. In all four ! cases B is positive deffinite. ! LA_SYGVD and LA_HEGVD use a divide and conquer algorithm. If ! eigenvectors are desired, they can be much faster than LA_SYGV and ! LA_HEGV for large matrices but use more workspace. ! ! ========= ! ! SUBROUTINE LA_SYGV / LA_SYGVD / LA_HEGV / LA_HEGVD( A, B, & ! W, ITYPE=itype, JOBZ=jobz, UPLO=uplo, INFO=info ) ! <type>(<wp>), INTENT(INOUT) :: A(:,:), B(:,:) ! REAL(<wp>), INTENT(OUT) :: W(:) ! INTEGER, INTENT(IN), OPTIONAL :: ITYPE ! CHARACTER(LEN=1), INTENT(IN), OPTIONAL :: JOBZ, UPLO ! INTEGER, INTENT(OUT), OPTIONAL :: INFO ! where ! <type> ::= REAL | COMPLEX ! <wp> ::= KIND(1.0) | KIND(1.0D0) ! ! Arguments ! ========= ! ! A (input/output) REAL or COMPLEX square array, shape (:,:). ! On entry, the matrix A. ! If UPLO = 'U', the upper triangular part of A contains the ! upper triangular part of matrix A. If UPLO = 'L', the lower ! triangular part of A contains the lower triangular part of ! matrix A. ! On exit, if JOBZ = 'V', then the columns of A contain the ! eigenvectors, normalized as follows: ! if ITYPE = 1 or 2: Z^H*B*Z = I , ! if ITYPE = 3: Z^H*B^-1*Z = I . ! If JOBZ = 'N', then the upper triangle (if UPLO = 'U') or the ! lower triangle (if UPLO = 'L') of A, including the diagonal, ! is destroyed. ! B (input/output) REAL or COMPLEX square array, shape (:,:) with ! size(B,1) = size(A,1). ! On entry, the matrix B. If UPLO = 'U', the upper triangular ! part of B contains the upper triangular part of matrix B. If ! UPLO = 'L', the lower triangular part of B contains the lower ! triangular part of matrix B. ! On exit, if the part of B containing the matrix is overwritten ! by the triangular factor U or L of the Cholesky factorization ! B = U^H*U or B = L*L^H , respectively. ! W (output) REAL array, shape (:) with size(W) = size(A,1). ! The eigenvalues in ascending order. ! ITYPE Optional (input) INTEGER. ! Specifies the problem type to be solved: ! = 1: A*z = lambda*B*z ! = 2: A*B*z = lambda*z ! = 3: B*A*z = lambda*z ! Default value: 1. ! JOBZ Optional (input) CHARACTER(LEN=1). ! = 'N': Compute eigenvalues only; ! = 'V': Compute eigenvalues and eigenvectors. ! Default value: 'N'. ! UPLO Optional (input) CHARACTER(LEN=1). ! = 'U': Upper triangles of A and B are stored; ! = 'L': Lower triangles of A and B are stored. ! Default value: 'U'. ! INFO Optional (output) INTEGER. ! = 0: successful exit. ! < 0: if INFO = -i, the i-th argument had an illegal value. ! > 0: the algorithm failed to converge or matrix B is not ! positive deffinite: ! <= n: if INFO = i, i off-diagonal elements of an ! intermediate tridiagonal form did not converge to ! zero. ! > n: if INFO = n+i, for 1 <= i <= n, then the leading minor ! of order i of B is not positive deffinite. The ! factorization of B could not be completed and no ! eigenvalues or eigenvectors were computed. ! n is the order of A. ! If INFO is not present and an error occurs, then the program is ! terminated with an error message. !------------------------------------------------------------------------ ! .. LOCAL PARAMETERS .. ! .. LOCAL SCALARS .. CHARACTER(LEN=1) :: LJOBZ, LUPLO ! .. LOCAL ARRAYS .. COMPLEX(WP), POINTER :: WORK(:) INTEGER, POINTER :: IWORK(:) COMPLEX(WP) :: WORKMIN(1) INTEGER :: IWORKMIN(1) ! .. INTRINSIC FUNCTIONS .. INTRINSIC SIZE, MAX, PRESENT ! .. EXECUTABLE STATEMENTS .. LINFO = 0; N = SIZE(A,1); LD = MAX(1,N); ISTAT = 0 IF( PRESENT(ITYPE) )THEN LITYPE = ITYPE ELSE LITYPE = 1 END IF IF( PRESENT(JOBZ) ) THEN LJOBZ = JOBZ ELSE LJOBZ = 'N' END IF IF( PRESENT(UPLO) ) THEN LUPLO = UPLO ELSE LUPLO = 'U' END IF ! .. TEST THE ARGUMENTS .. IF( SIZE( A, 2 ) /= N .OR. N < 0 )THEN LINFO = -1 ELSE IF( SIZE( B, 1 ) /= N .OR. SIZE( B, 2 ) /= N )THEN LINFO = -2 ELSE IF( SIZE( W ) /= N )THEN LINFO = -3 ELSE IF( LITYPE < 1 .OR. LITYPE > 3 )THEN LINFO = -4 ELSE IF( .NOT.LSAME(LJOBZ,'N') .AND. .NOT.LSAME(LJOBZ,'V') )THEN LINFO = -5 ELSE IF( .NOT.LSAME(LUPLO,'U') .AND. .NOT.LSAME(LUPLO,'L') )THEN LINFO = -6 ELSE IF( N > 0 )THEN ! .. DETERMINE THE WORKSPACE .. ! .. QUERING THE SIZE OF WORKSPACE .. ENDIF CALL ERINFO(LINFO, SRNAME, INFO, ISTAT)
{ "alphanum_fraction": 0.5439239873, "author": null, "avg_line_length": 40.5337837838, "converted": null, "ext": "f90", "file": null, "hexsha": "f06960500fd06d00e7c57a75bfb2690b2f41cd58", "include": null, "lang": "FORTRAN", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 7, "max_forks_repo_forks_event_max_datetime": "2021-04-15T07:12:40.000Z", "max_forks_repo_forks_event_min_datetime": "2018-12-29T15:34:01.000Z", "max_forks_repo_head_hexsha": "bcd9d4b706f4213a6a4c0ebb4521754ffeff3752", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "MattBurn/LAPACK95", "max_forks_repo_path": "src/la_csygvd.f90", "max_issues_count": 1, "max_issues_repo_head_hexsha": "bcd9d4b706f4213a6a4c0ebb4521754ffeff3752", "max_issues_repo_issues_event_max_datetime": "2018-12-31T06:45:11.000Z", "max_issues_repo_issues_event_min_datetime": "2018-12-30T15:38:47.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "MattBurn/LAPACK95", "max_issues_repo_path": "src/la_csygvd.f90", "max_line_length": 74, "max_stars_count": 8, "max_stars_repo_head_hexsha": "bcd9d4b706f4213a6a4c0ebb4521754ffeff3752", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "MattBurn/LAPACK95", "max_stars_repo_path": "src/la_csygvd.f90", "max_stars_repo_stars_event_max_datetime": "2022-03-02T10:09:22.000Z", "max_stars_repo_stars_event_min_datetime": "2018-12-29T15:07:54.000Z", "num_tokens": 1792, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 5999 }
# This contains inherited classes from detectron2, which were defined wit h he detectron2 version on January 21, 2019. # the changes are made to allow for logging validation set metrics during training and to allow for a custom data loader for tif imagery. import copy import logging import numpy as np import os import operator import torch.utils.data import time from detectron2.utils.comm import get_world_size from detectron2.utils.env import seed_all_rng from detectron2.utils.logger import log_first_n from detectron2.data import samplers from detectron2.data.common import ( AspectRatioGroupedDataset, DatasetFromList, MapDataset, ) from detectron2.data.dataset_mapper import DatasetMapper import detectron2.data.detection_utils as utils import detectron2.utils.comm as comm from detectron2.data.build import * from detectron2.data.build import ( worker_init_reset_seed, trivial_batch_collator ) # it's hidden by _all_ in build.py :( from detectron2.data import transforms as T from detectron2.engine import DefaultTrainer, hooks from detectron2.evaluation import COCOEvaluator, DatasetEvaluators import torch def tif_dataset_mapper(dataset_dict): # Implement a mapper, similar to the default DatasetMapper, but with your own customizations # it will be modified by code below dataset_dict = copy.deepcopy(dataset_dict) image = utils.read_image( dataset_dict["file_name"] ) # relies on my own edit that updates the func to be able to read tif imagery with imageio # image, transforms = T.apply_transform_gens([T.Resize((800, 800))], image) dataset_dict["image"] = torch.as_tensor( image.transpose(2, 0, 1).astype("float32")) # annos = [ # utils.transform_instance_annotations(obj, transforms, image.shape[:2]) # for obj in dataset_dict.pop("annotations") # if obj.get("iscrowd", 0) == 0 # ] annos = [ obj for obj in dataset_dict.pop("annotations") if obj.get("iscrowd", 0) == 0 ] instances = utils.annotations_to_instances(annos, image.shape[:2]) dataset_dict["instances"] = utils.filter_empty_instances(instances) return dataset_dict def build_detection_validation_loader(cfg, mapper=None): """ A data loader is created by the following steps: 1. Use the dataset names in config to query :class:`DatasetCatalog`, and obtain a list of dicts. 2. Start workers to work on the dicts. Each worker will: * Map each metadata dict into another format to be consumed by the model. * Batch them by simply putting dicts into a list. The batched ``list[mapped_dict]`` is what this dataloader will return. Args: cfg (CfgNode): the config mapper (callable): a callable which takes a sample (dict) from dataset and returns the format to be consumed by the model. By default it will be `DatasetMapper(cfg, True)`. Returns: an infinite iterator of validation data """ num_workers = get_world_size() images_per_batch = cfg.SOLVER.IMS_PER_BATCH assert ( images_per_batch % num_workers == 0 ), "SOLVER.IMS_PER_BATCH ({}) must be divisible by the number of workers ({}).".format( images_per_batch, num_workers ) assert ( images_per_batch >= num_workers ), "SOLVER.IMS_PER_BATCH ({}) must be larger than the number of workers ({}).".format( images_per_batch, num_workers ) images_per_worker = images_per_batch // num_workers dataset_dicts = get_detection_dataset_dicts( # this is the only difference between the valdiation loader and train loader. cfg.DATASETS.VALIDATION, filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS, min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE if cfg.MODEL.KEYPOINT_ON else 0, proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN if cfg.MODEL.LOAD_PROPOSALS # won't use this I think but if I do, needs to be changed for valdiation set. else None, ) dataset = DatasetFromList(dataset_dicts, copy=False) if mapper is None: mapper = DatasetMapper(cfg, True) dataset = MapDataset(dataset, mapper) sampler_name = ( cfg.DATALOADER.SAMPLER_TRAIN ) # valdiation should use same batching and sampling scheme as training (rather than batch of size 1) logger = logging.getLogger(__name__) logger.info("Using training sampler {}".format(sampler_name)) if sampler_name == "TrainingSampler": sampler = samplers.TrainingSampler(len(dataset)) elif sampler_name == "RepeatFactorTrainingSampler": sampler = samplers.RepeatFactorTrainingSampler( dataset_dicts, cfg.DATALOADER.REPEAT_THRESHOLD ) else: raise ValueError("Unknown training sampler: {}".format(sampler_name)) if cfg.DATALOADER.ASPECT_RATIO_GROUPING: data_loader = torch.utils.data.DataLoader( dataset, sampler=sampler, num_workers=cfg.DATALOADER.NUM_WORKERS, batch_sampler=None, collate_fn=operator.itemgetter( 0 ), # don't batch, but yield individual elements worker_init_fn=worker_init_reset_seed, ) # yield individual mapped dict data_loader = AspectRatioGroupedDataset(data_loader, images_per_worker) else: batch_sampler = torch.utils.data.sampler.BatchSampler( sampler, images_per_worker, drop_last=True ) # drop_last so the batch always have the same size data_loader = torch.utils.data.DataLoader( dataset, num_workers=cfg.DATALOADER.NUM_WORKERS, batch_sampler=batch_sampler, collate_fn=trivial_batch_collator, worker_init_fn=worker_init_reset_seed, ) return data_loader def run_validation_step(self): """ Added to write out validation metrics. Used as a CallbackHook in the after_step stage in Trainer.build_hooks(). """ if (self.iter % self.cfg.VALIDATION_PERIOD) == 0: validation_data = next(self.validation_data_loader_iter) val_losses_dict = self.model(validation_data) val_losses = sum(loss for loss in val_losses_dict.values()) self._detect_anomaly(val_losses, val_losses_dict) val_metrics_dict = val_losses_dict # val_metrics_dict["data_time"] = data_time self._write_validation_metrics(val_metrics_dict) class Trainer(DefaultTrainer): """ This subclasses the Default Trainer to plot a validation loss curve alongside train loss curve in tensorboard. This requires a train/validation/test split upfront. """ def __init__(self, cfg): """ Args: cfg (CfgNode): """ super().__init__(cfg) validation_data_loader = self.build_validation_loader(cfg) self.validation_data_loader_iter = iter(validation_data_loader) @classmethod def build_evaluator(cls, cfg, dataset_name): output_folder = os.path.join(cfg.OUTPUT_DIR, "inference") evaluators = [COCOEvaluator(dataset_name, cfg, True, output_folder)] return DatasetEvaluators(evaluators) @classmethod def build_test_loader(cls, cfg, dataset_name): return build_detection_test_loader(cfg, dataset_name, mapper=tif_dataset_mapper) @classmethod def build_validation_loader(cls, cfg): return build_detection_validation_loader(cfg, mapper=tif_dataset_mapper) @classmethod def build_train_loader(cls, cfg): return build_detection_train_loader(cfg, mapper=tif_dataset_mapper) def run_step(self): """ Implement the standard training logic described above. """ assert self.model.training, "[SimpleTrainer] model was changed to eval mode!" start = time.perf_counter() """ If your want to do something with the data, you can wrap the dataloader. """ data = next(self._data_loader_iter) data_time = time.perf_counter() - start """ If your want to do something with the losses, you can wrap the model. """ loss_dict = self.model(data) losses = sum(loss for loss in loss_dict.values()) self._detect_anomaly(losses, loss_dict) metrics_dict = loss_dict metrics_dict["data_time"] = data_time self._write_metrics(metrics_dict) validation_data = next(self.validation_data_loader_iter) val_losses_dict = self.model(validation_data) val_losses = sum(loss for loss in val_losses_dict.values()) self._detect_anomaly(val_losses, val_losses_dict) val_metrics_dict = val_losses_dict val_metrics_dict["data_time"] = data_time self._write_validation_metrics(val_metrics_dict) """ If you need accumulate gradients or something similar, you can wrap the optimizer with your custom `zero_grad()` method. """ self.optimizer.zero_grad() losses.backward() """ If you need gradient clipping/scaling or other processing, you can wrap the optimizer with your custom `step()` method. """ self.optimizer.step() def _write_validation_metrics(self, metrics_dict: dict): """ Args: metrics_dict (dict): dict of scalar metrics """ metrics_dict = { k: v.detach().cpu().item() if isinstance(v, torch.Tensor) else float(v) for k, v in metrics_dict.items() } # gather metrics among all workers for logging # This assumes we do DDP-style training, which is currently the only # supported method in detectron2. all_metrics_dict = comm.gather(metrics_dict) if comm.is_main_process(): if "data_time" in all_metrics_dict[0]: # data_time among workers can have high variance. The actual latency # caused by data_time is the maximum among workers. data_time = np.max([x.pop("data_time") for x in all_metrics_dict]) self.storage.put_scalar("data_time", data_time) # average the rest metrics. val added to metric key names # compared to original _write_metrics func in train_loop.py metrics_dict = { "val_"+k: np.mean([x[k] for x in all_metrics_dict]) for k in all_metrics_dict[0].keys() } total_losses_reduced = sum(loss for loss in metrics_dict.values()) self.storage.put_scalar("total_loss/val", total_losses_reduced) if len(metrics_dict) > 1: self.storage.put_scalars(**metrics_dict) def _write_metrics(self, metrics_dict: dict): """ Args: metrics_dict (dict): dict of scalar metrics """ metrics_dict = { k: v.detach().cpu().item() if isinstance(v, torch.Tensor) else float(v) for k, v in metrics_dict.items() } # gather metrics among all workers for logging # This assumes we do DDP-style training, which is currently the only # supported method in detectron2. all_metrics_dict = comm.gather(metrics_dict) if comm.is_main_process(): if "data_time" in all_metrics_dict[0]: # data_time among workers can have high variance. The actual latency # caused by data_time is the maximum among workers. data_time = np.max([x.pop("data_time") for x in all_metrics_dict]) self.storage.put_scalar("data_time", data_time) # average the rest metrics metrics_dict = { "train_"+k: np.mean([x[k] for x in all_metrics_dict]) for k in all_metrics_dict[0].keys() } total_losses_reduced = sum(loss for loss in metrics_dict.values()) self.storage.put_scalar("total_loss/train", total_losses_reduced) if len(metrics_dict) > 1: self.storage.put_scalars(**metrics_dict)
{ "alphanum_fraction": 0.6674586269, "author": null, "avg_line_length": 39.3741935484, "converted": null, "ext": "py", "file": null, "hexsha": "bb194a4b4af081a7c8b6a84ec28706d6772b9301", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 3, "max_forks_repo_forks_event_max_datetime": "2021-12-01T15:52:06.000Z", "max_forks_repo_forks_event_min_datetime": "2018-11-19T23:02:01.000Z", "max_forks_repo_head_hexsha": "4657ed1d103acb37dc974aa6af2f0d3a3398e987", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "ecohydro/CropMask_RCNN", "max_forks_repo_path": "cropmask/detectron2_reclass.py", "max_issues_count": 32, "max_issues_repo_head_hexsha": "4657ed1d103acb37dc974aa6af2f0d3a3398e987", "max_issues_repo_issues_event_max_datetime": "2020-12-31T19:48:41.000Z", "max_issues_repo_issues_event_min_datetime": "2019-02-21T21:14:18.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "ecohydro/CropMask_RCNN", "max_issues_repo_path": "cropmask/detectron2_reclass.py", "max_line_length": 137, "max_stars_count": 13, "max_stars_repo_head_hexsha": "4657ed1d103acb37dc974aa6af2f0d3a3398e987", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "ecohydro/CropMask_RCNN", "max_stars_repo_path": "cropmask/detectron2_reclass.py", "max_stars_repo_stars_event_max_datetime": "2021-07-12T06:28:31.000Z", "max_stars_repo_stars_event_min_datetime": "2019-03-01T23:41:27.000Z", "num_tokens": 2659, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 12206 }
import numpy as np from ssr.ext.plyfile import PlyData, PlyElement from ssr.utility.logging_extension import logger from ssr.ssr_types.point import Point, Measurement class Face: # Do NOT confuse this class with a real triangle. # A face saves ONLY VERTEX INDICES, NO 3D information. def __init__(self, initial_indices=np.array([-1, -1, -1], dtype=int)): # We do NOT DEFINE COLOR PER FACE, since the color of the face is # determined by the corresponding vertex colors self.vertex_indices = np.array(initial_indices, dtype=int) def __str__(self): return str(self.vertex_indices) class PLYFileHandler: @staticmethod def __ply_data_vertices_to_vetex_list(ply_data): vertex_data_type_names = ply_data["vertex"].data.dtype.names use_color = False if ( "red" in vertex_data_type_names and "green" in vertex_data_type_names and "blue" in vertex_data_type_names ): use_color = True vertices = [] value_keys = [ x for x, y in sorted( ply_data["vertex"].data.dtype.fields.items(), key=lambda k: k[1], ) ] non_scalar_value_keys = [ "x", "y", "z", "red", "green", "blue", "nx", "ny", "nz", "measurements", ] scalar_value_keys = [ value_key for value_key in value_keys if not value_key in non_scalar_value_keys ] logger.info( "Found the following vertex properties: " + str(value_keys) ) logger.info("Found " + str(len(ply_data["vertex"].data)) + " vertices") for point_index, line in enumerate(ply_data["vertex"].data): current_point = Point() current_point.coord = np.array([line["x"], line["y"], line["z"]]) if use_color: current_point.color = np.array( [line["red"], line["green"], line["blue"]] ) current_point.id = point_index for scalar_value_key in scalar_value_keys: current_point.scalars[scalar_value_key] = line[ scalar_value_key ] if "measurements" in line.dtype.names: elements_per_measurement = 4 current_point.measurements = [] for measurement_idx in range( len(line["measurements"]) / elements_per_measurement ): array_idx = measurement_idx * elements_per_measurement slice = line["measurements"][ array_idx : array_idx + elements_per_measurement ] current_point.measurements.append( Measurement.init_from_list(slice) ) vertices.append(current_point) ply_data_vertex_dtype = ply_data["vertex"].dtype ply_data_vertex_data_dtype = ply_data["vertex"].data.dtype return vertices, ply_data_vertex_dtype, ply_data_vertex_data_dtype @staticmethod def __ply_data_faces_to_face_list(ply_data): faces = [] ply_data_face_type = None ply_data_face_data_type = None if "face" in ply_data: # read faces ply_data_face_type = ply_data["face"].dtype logger.info("Found " + str(len(ply_data["face"].data)) + " faces") for line in ply_data["face"].data["vertex_indices"]: current_face = Face() current_face.vertex_indices = np.array( [line[0], line[1], line[2]] ) faces.append(current_face) ply_data_face_data_type = [("vertex_indices", "i4", (3,))] face_names = ply_data["face"].data.dtype.names if ( "red" in face_names and "green" in face_names and "blue" in face_names ): ply_data_face_data_type = [ ("vertex_indices", "i4", (3,)), ("red", "u1"), ("green", "u1"), ("blue", "u1"), ] return faces, ply_data_face_type, ply_data_face_data_type @staticmethod def __vertices_to_ply_vertex_element( point_list, ply_data_vertex_data_dtype_list ): ply_data_vertex_data_dtype = np.dtype(ply_data_vertex_data_dtype_list) # if measurements are used, then we do not know one dimension of the array vertex_output_array = np.empty( (len(point_list),), dtype=ply_data_vertex_data_dtype ) with_color = False if ( "red" in ply_data_vertex_data_dtype.names and "green" in ply_data_vertex_data_dtype.names and "blue" in ply_data_vertex_data_dtype.names ): with_color = True with_normals = False if ( "nx" in ply_data_vertex_data_dtype.names and "ny" in ply_data_vertex_data_dtype.names and "nz" in ply_data_vertex_data_dtype.names ): with_normals = True with_measurements = "measurements" in ply_data_vertex_data_dtype.names # set all the values, offered / defined by property_type_list for index, point in enumerate(point_list): # row = np.empty(1, dtype=ply_data_vertex_data_dtype) vertex_output_array[index]["x"] = point.coord[0] vertex_output_array[index]["y"] = point.coord[1] vertex_output_array[index]["z"] = point.coord[2] if with_color: vertex_output_array[index]["red"] = point.color[0] vertex_output_array[index]["green"] = point.color[1] vertex_output_array[index]["blue"] = point.color[2] if with_normals: vertex_output_array[index]["nx"] = point.normal[0] vertex_output_array[index]["ny"] = point.normal[1] vertex_output_array[index]["nz"] = point.normal[2] for scalar_key in point.scalars: vertex_output_array[index][scalar_key] = point.scalars[ scalar_key ] if with_measurements: measurements = [] for measurement in point.measurements: measurements += measurement.to_list() vertex_output_array[index]["measurements"] = measurements description = PlyElement.describe( vertex_output_array, name="vertex", # possible values for val_types # ['int8', 'i1', 'char', 'uint8', 'u1', 'uchar', 'b1', # 'int16', 'i2', 'short', 'uint16', 'u2', 'ushort', # 'int32', 'i4', 'int', 'uint32', 'u4', 'uint', # 'float32', 'f4', 'float', 'float64', 'f8', 'double'] val_types={"measurements": "float"}, ) return description @staticmethod def __faces_to_ply_face_element(face_list, property_type_list): face_output_array = np.empty(len(face_list), dtype=property_type_list) for index, face in enumerate(face_list): row = np.empty(1, dtype=property_type_list) # We don't use face colors, the color of the faces is defined using # the vertex colors! row[ "vertex_indices" ] = face.vertex_indices # face.vertex_indices is a np.array face_output_array[index] = row output_ply_data_face_element = PlyElement.describe( face_output_array, "face" ) return output_ply_data_face_element @staticmethod def __cameras_2_ply_vertex_element(camera_list, property_type_list): camera_output_array = np.empty( len(camera_list), dtype=property_type_list ) for index, camera in enumerate(camera_list): row = np.empty(1, dtype=property_type_list) row["x"] = camera.get_camera_center()[0] row["y"] = camera.get_camera_center()[1] row["z"] = camera.get_camera_center()[2] row["red"] = camera.color[0] row["green"] = camera.color[1] row["blue"] = camera.color[2] row["nx"] = camera.normal[0] row["ny"] = camera.normal[1] row["nz"] = camera.normal[2] camera_output_array[index] = row return PlyElement.describe(camera_output_array, "vertex") @staticmethod def parse_ply_file_extended(ifp): logger.info("Parse PLY File: ...") ply_data = PlyData.read(ifp) ( vertices, ply_data_vertex_dtype, ply_data_vertex_data_dtype, ) = PLYFileHandler.__ply_data_vertices_to_vetex_list(ply_data) ( faces, ply_data_face_type, ply_data_face_data_type, ) = PLYFileHandler.__ply_data_faces_to_face_list(ply_data) logger.info("Parse PLY File: Done") # return always 6 arguments. However, the latter may be empty return ( vertices, ply_data_vertex_dtype, ply_data_vertex_data_dtype, faces, ply_data_face_type, ply_data_face_data_type, ) @staticmethod def parse_ply_file(ifp): logger.info("Parse PLY File: ...") logger.vinfo("ifp", ifp) ply_data = PlyData.read(ifp) vertices, _, _ = PLYFileHandler.__ply_data_vertices_to_vetex_list( ply_data ) faces, _, _ = PLYFileHandler.__ply_data_faces_to_face_list(ply_data) logger.info("Parse PLY File: Done") return vertices, faces @staticmethod def write_ply_file_from_vertex_mat(ofp, vertex_mat): vertices = [] for entry in vertex_mat: vertices.append(Point(coord=entry)) PLYFileHandler.write_ply_file(ofp, vertices) @staticmethod def build_type_list( vertices, with_colors, with_normals, with_measurements ): ply_data_vertex_data_dtype_list = [ ("x", "<f4"), ("y", "<f4"), ("z", "<f4"), ] if with_colors: ply_data_vertex_data_dtype_list += [ ("red", "u1"), ("green", "u1"), ("blue", "u1"), ] if with_normals: ply_data_vertex_data_dtype_list += [ ("nx", "<f4"), ("ny", "<f4"), ("nz", "<f4"), ] if len(vertices) > 0: for scalar_keys in vertices[0].scalars: ply_data_vertex_data_dtype_list += [(scalar_keys, "<f4")] if with_measurements: # since the length of the measurements varies, we use an object data type here ply_data_vertex_data_dtype_list += [("measurements", object)] return ply_data_vertex_data_dtype_list @staticmethod def write_ply_file( ofp, vertices, with_colors=True, with_normals=False, faces=None, plain_text_output=False, with_measurements=False, ): logger.info("write_ply_file: " + ofp) ply_data_vertex_data_dtype_list = PLYFileHandler.build_type_list( vertices, with_colors, with_normals, with_measurements ) logger.vinfo( "ply_data_vertex_data_dtype_list", ply_data_vertex_data_dtype_list ) output_ply_data_vertex_element = ( PLYFileHandler.__vertices_to_ply_vertex_element( vertices, ply_data_vertex_data_dtype_list ) ) if faces is None or len(faces) == 0: logger.info("Write File With Vertices Only (no faces)") output_data = PlyData( [output_ply_data_vertex_element], text=plain_text_output ) else: logger.info("Write File With Faces") logger.info("Number faces" + str(len(faces))) ply_data_face_data_type = [("vertex_indices", "i4", (3,))] # we do not define colors for faces, # since we use the vertex colors to colorize the face output_ply_data_face_element = ( PLYFileHandler.__faces_to_ply_face_element( faces, ply_data_face_data_type ) ) output_data = PlyData( [output_ply_data_vertex_element, output_ply_data_face_element], text=plain_text_output, ) output_data.write(ofp) @staticmethod def write_camera_ply_file(ofp, cameras, plain_text_output=True): ply_data_vertex_data_dtype_list = [ ("x", "<f4"), ("y", "<f4"), ("z", "<f4"), ] ply_data_vertex_data_dtype_list += [ ("red", "u1"), ("green", "u1"), ("blue", "u1"), ] ply_data_vertex_data_dtype_list += [ ("nx", "<f4"), ("ny", "<f4"), ("nz", "<f4"), ] ply_data_vertex_data_dtype = np.dtype(ply_data_vertex_data_dtype_list) output_ply_data_vertex_element = ( PLYFileHandler.__cameras_2_ply_vertex_element( cameras, ply_data_vertex_data_dtype ) ) # [('x', '<f4'), ('y', '<f4'), ('z', '<f4'), ('red', 'u1'), ('green', 'u1'), ('blue', 'u1')] logger.info("Write (Camera) File With Vertices Only (no faces)") output_data = PlyData( [output_ply_data_vertex_element], text=plain_text_output ) output_data.write(ofp) @staticmethod def read_and_write_test(ifp, ofp): vertices, faces = PLYFileHandler.parse_ply_file(ifp) PLYFileHandler.write_ply_file(ofp, vertices, faces=faces) if __name__ == "__main__": ply_ofp = "out.ply" measurements1 = [ Measurement(1222.428, 2, 3, 4), Measurement(8, 9, 10.1, 11.2), Measurement(28, 29, 30, 31), Measurement(12213, 29, 30, 31), ] measurements2 = [Measurement(11, 12, 13, 14), Measurement(18, 19, 20, 21)] p_1 = Point([0, 0, 0]) p_1.measurements = measurements1 p_2 = Point([0, 0, 1]) p_2.measurements = measurements1 p_3 = Point([0, 1, 1]) p_3.measurements = measurements2 p_4 = Point([1, 0, 0]) p_4.measurements = measurements2 p_5 = Point([1, 1, 1]) p_5.measurements = measurements2 points = [p_1, p_2, p_3, p_4, p_5] PLYFileHandler.write_ply_file( ply_ofp, points, plain_text_output=True, with_measurements=True ) vertices, faces = PLYFileHandler.parse_ply_file(ply_ofp) p_1, p_2, p_3, p_4, p_5 = vertices for p in vertices: logger.info(p.coord, p.color, p.measurements)
{ "alphanum_fraction": 0.5634531229, "author": null, "avg_line_length": 33.0021881838, "converted": null, "ext": "py", "file": null, "hexsha": "ad78c1a9c9506221724b8eddba8c38c68f33a474", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 6, "max_forks_repo_forks_event_max_datetime": "2022-03-18T14:09:16.000Z", "max_forks_repo_forks_event_min_datetime": "2021-05-11T12:18:07.000Z", "max_forks_repo_head_hexsha": "24aa67ace67e69568cbc7e01a9e8b407463366f4", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "jieeeeeeeeeee/SatelliteSurfaceReconstruction", "max_forks_repo_path": "ssr/file_handler/ply_file_handler.py", "max_issues_count": 3, "max_issues_repo_head_hexsha": "24aa67ace67e69568cbc7e01a9e8b407463366f4", "max_issues_repo_issues_event_max_datetime": "2021-05-25T08:40:13.000Z", "max_issues_repo_issues_event_min_datetime": "2021-04-06T14:01:43.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "jieeeeeeeeeee/SatelliteSurfaceReconstruction", "max_issues_repo_path": "ssr/file_handler/ply_file_handler.py", "max_line_length": 100, "max_stars_count": 34, "max_stars_repo_head_hexsha": "7127bc0fb36155e31ae2928c18a65d562d7d29ba", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "SBCV/SatelliteSurfaceReconstruction", "max_stars_repo_path": "ssr/file_handler/ply_file_handler.py", "max_stars_repo_stars_event_max_datetime": "2022-03-27T14:12:34.000Z", "max_stars_repo_stars_event_min_datetime": "2021-04-03T13:12:34.000Z", "num_tokens": 3447, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 15082 }
// Copyright (c) 2007-2017 Hartmut Kaiser // Copyright (c) 2008-2009 Chirag Dekate, Anshul Tandon // Copyright (c) 2012-2013 Thomas Heller // // Distributed under the Boost Software License, Version 1.0. (See accompanying // file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) #include <hpx/runtime/threads/topology.hpp> #include <hpx/compat/thread.hpp> #include <hpx/error_code.hpp> #include <hpx/exception.hpp> #include <hpx/throw_exception.hpp> #include <hpx/util/assert.hpp> #include <hpx/util/format.hpp> #include <hpx/util/logging.hpp> #include <hpx/util/spinlock.hpp> #include <hpx/runtime.hpp> #include <hpx/runtime/naming/address.hpp> #include <hpx/runtime/threads/cpu_mask.hpp> #include <hpx/runtime/threads/topology.hpp> #include <boost/io/ios_state.hpp> #include <boost/scoped_ptr.hpp> #include <cstddef> #include <iomanip> #include <iostream> #include <mutex> #include <string> #include <vector> #include <memory> #include <errno.h> #include <hwloc.h> #if HWLOC_API_VERSION < 0x00010b00 # define HWLOC_OBJ_NUMANODE HWLOC_OBJ_NODE #endif #if defined(__ANDROID__) && defined(ANDROID) #include <cpu-features.h> #endif #if defined(__bgq__) #include <hwi/include/bqc/A2_inlines.h> #endif #if defined(_POSIX_VERSION) #include <sys/syscall.h> #include <sys/resource.h> #endif namespace hpx { namespace threads { namespace detail { std::size_t hwloc_hardware_concurrency() { threads::topology& top = threads::create_topology(); return top.get_number_of_pus(); } void write_to_log(char const* valuename, std::size_t value) { LTM_(debug) << "topology: " << valuename << ": " << value; //-V128 } void write_to_log_mask(char const* valuename, mask_cref_type value) { LTM_(debug) << "topology: " << valuename << ": " HPX_CPU_MASK_PREFIX << std::hex << value; } void write_to_log(char const* valuename, std::vector<std::size_t> const& values) { LTM_(debug) << "topology: " << valuename << "s, size: " //-V128 << values.size(); std::size_t i = 0; for (std::size_t value : values) { LTM_(debug) << "topology: " << valuename //-V128 << "(" << i++ << "): " << value; } } void write_to_log_mask(char const* valuename, std::vector<mask_type> const& values) { LTM_(debug) << "topology: " << valuename << "s, size: " //-V128 << values.size(); std::size_t i = 0; for (mask_cref_type value : values) { LTM_(debug) << "topology: " << valuename //-V128 << "(" << i++ << "): " HPX_CPU_MASK_PREFIX << std::hex << value; } } std::size_t get_index(hwloc_obj_t obj) { // on Windows logical_index is always -1 if (obj->logical_index == ~0x0u) return static_cast<std::size_t>(obj->os_index); return static_cast<std::size_t>(obj->logical_index); } hwloc_obj_t adjust_node_obj(hwloc_obj_t node) noexcept { #if HWLOC_API_VERSION >= 0x00020000 // www.open-mpi.org/projects/hwloc/doc/hwloc-v2.0.0-letter.pdf: // Starting with hwloc v2.0, NUMA nodes are not in the main tree // anymore. They are attached under objects as Memory Children // on the side of normal children. while (hwloc_obj_type_is_memory(node->type)) node = node->parent; HPX_ASSERT(node); #endif return node; } }}} namespace hpx { namespace threads { /////////////////////////////////////////////////////////////////////////// std::ostream& operator<<(std::ostream& os, hpx_hwloc_bitmap_wrapper const* bmp) { char buffer[256]; hwloc_bitmap_snprintf(buffer, 256, bmp->bmp_); os << buffer; return os; } /////////////////////////////////////////////////////////////////////////// mask_type topology::get_service_affinity_mask( mask_cref_type used_processing_units, error_code& ec) const { // We bind the service threads to the first NUMA domain. This is useful // as the first NUMA domain is likely to have the PCI controllers etc. mask_cref_type machine_mask = this->get_numa_node_affinity_mask(0, ec); if (ec || !any(machine_mask)) return mask_type(); if (&ec != &throws) ec = make_success_code(); mask_type res = ~used_processing_units & machine_mask; return (!any(res)) ? machine_mask : res; } bool topology::reduce_thread_priority(error_code& ec) const { #ifdef HPX_HAVE_NICE_THREADLEVEL #if defined(__linux__) && !defined(__ANDROID__) && !defined(__bgq__) pid_t tid; tid = syscall(SYS_gettid); if (setpriority(PRIO_PROCESS, tid, 19)) { HPX_THROWS_IF(ec, no_success, "topology::reduce_thread_priority", "setpriority returned an error"); return false; } #elif defined(WIN32) || defined(_WIN32) || defined(__WIN32__) if (!SetThreadPriority(GetCurrentThread(), THREAD_PRIORITY_LOWEST)) { HPX_THROWS_IF(ec, no_success, "topology::reduce_thread_priority", "SetThreadPriority returned an error"); return false; } #elif defined(__bgq__) ThreadPriority_Low(); #endif #endif return true; } /////////////////////////////////////////////////////////////////////////// mask_type topology::empty_mask = mask_type(); topology::topology() : topo(nullptr), machine_affinity_mask_(0) { // {{{ int err = hwloc_topology_init(&topo); if (err != 0) { HPX_THROW_EXCEPTION(no_success, "topology::topology", "Failed to init hwloc topology"); } err = hwloc_topology_load(topo); if (err != 0) { HPX_THROW_EXCEPTION(no_success, "topology::topology", "Failed to load hwloc topology"); } init_num_of_pus(); socket_numbers_.reserve(num_of_pus_); numa_node_numbers_.reserve(num_of_pus_); core_numbers_.reserve(num_of_pus_); // Initialize each set of data entirely, as some of the initialization // routines rely on access to other pieces of topology data. The // compiler will optimize the loops where possible anyways. std::size_t num_of_sockets = get_number_of_sockets(); if (num_of_sockets == 0) num_of_sockets = 1; for (std::size_t i = 0; i < num_of_pus_; ++i) { std::size_t socket = init_socket_number(i); HPX_ASSERT(socket < num_of_sockets); socket_numbers_.push_back(socket); } std::size_t num_of_nodes = get_number_of_numa_nodes(); if (num_of_nodes == 0) num_of_nodes = 1; for (std::size_t i = 0; i < num_of_pus_; ++i) { std::size_t numa_node = init_numa_node_number(i); HPX_ASSERT(numa_node < num_of_nodes); numa_node_numbers_.push_back(numa_node); } std::size_t num_of_cores = get_number_of_cores(); if (num_of_cores == 0) num_of_cores = 1; for (std::size_t i = 0; i < num_of_pus_; ++i) { std::size_t core_number = init_core_number(i); HPX_ASSERT(core_number < num_of_cores); core_numbers_.push_back(core_number); } machine_affinity_mask_ = init_machine_affinity_mask(); socket_affinity_masks_.reserve(num_of_pus_); numa_node_affinity_masks_.reserve(num_of_pus_); core_affinity_masks_.reserve(num_of_pus_); thread_affinity_masks_.reserve(num_of_pus_); for (std::size_t i = 0; i < num_of_pus_; ++i) { socket_affinity_masks_.push_back(init_socket_affinity_mask(i)); } for (std::size_t i = 0; i < num_of_pus_; ++i) { numa_node_affinity_masks_.push_back(init_numa_node_affinity_mask(i)); } for (std::size_t i = 0; i < num_of_pus_; ++i) { core_affinity_masks_.push_back(init_core_affinity_mask(i)); } for (std::size_t i = 0; i < num_of_pus_; ++i) { thread_affinity_masks_.push_back(init_thread_affinity_mask(i)); } } // }}} void topology::write_to_log() const { std::size_t num_of_sockets = get_number_of_sockets(); if (num_of_sockets == 0) num_of_sockets = 1; detail::write_to_log("num_sockets", num_of_sockets); std::size_t num_of_nodes = get_number_of_numa_nodes(); if (num_of_nodes == 0) num_of_nodes = 1; detail::write_to_log("num_of_nodes", num_of_nodes); std::size_t num_of_cores = get_number_of_cores(); if (num_of_cores == 0) num_of_cores = 1; detail::write_to_log("num_of_cores", num_of_cores); detail::write_to_log("num_of_pus", num_of_pus_); detail::write_to_log("socket_number", socket_numbers_); detail::write_to_log("numa_node_number", numa_node_numbers_); detail::write_to_log("core_number", core_numbers_); detail::write_to_log_mask("machine_affinity_mask", machine_affinity_mask_); detail::write_to_log_mask("socket_affinity_mask", socket_affinity_masks_); detail::write_to_log_mask("numa_node_affinity_mask", numa_node_affinity_masks_); detail::write_to_log_mask("core_affinity_mask", core_affinity_masks_); detail::write_to_log_mask("thread_affinity_mask", thread_affinity_masks_); } topology::~topology() { if (topo) hwloc_topology_destroy(topo); } std::size_t topology::get_pu_number( std::size_t num_core , std::size_t num_pu , error_code& ec ) const { // {{{ std::unique_lock<hpx::util::spinlock> lk(topo_mtx); int num_cores = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_CORE); // If num_cores is smaller 0, we have an error, it should never be zero // either to avoid division by zero, we should always have at least one // core if(num_cores <= 0) { HPX_THROWS_IF(ec, no_success, "topology::hwloc_get_nobjs_by_type", "Failed to get number of cores"); return std::size_t(-1); } num_core %= num_cores; //-V101 //-V104 //-V107 hwloc_obj_t core_obj; core_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_CORE, static_cast<unsigned>(num_core)); num_pu %= core_obj->arity; //-V101 //-V104 return std::size_t(core_obj->children[num_pu]->logical_index); } // }}} /////////////////////////////////////////////////////////////////////////// mask_cref_type topology::get_machine_affinity_mask( error_code& ec ) const { if (&ec != &throws) ec = make_success_code(); return machine_affinity_mask_; } mask_cref_type topology::get_socket_affinity_mask( std::size_t num_thread , error_code& ec ) const { // {{{ std::size_t num_pu = num_thread % num_of_pus_; if (num_pu < socket_affinity_masks_.size()) { if (&ec != &throws) ec = make_success_code(); return socket_affinity_masks_[num_pu]; } HPX_THROWS_IF(ec, bad_parameter , "hpx::threads::topology::get_socket_affinity_mask" , hpx::util::format( "thread number %1% is out of range", num_thread)); return empty_mask; } // }}} mask_cref_type topology::get_numa_node_affinity_mask( std::size_t num_thread , error_code& ec ) const { // {{{ std::size_t num_pu = num_thread % num_of_pus_; if (num_pu < numa_node_affinity_masks_.size()) { if (&ec != &throws) ec = make_success_code(); return numa_node_affinity_masks_[num_pu]; } HPX_THROWS_IF(ec, bad_parameter , "hpx::threads::topology::get_numa_node_affinity_mask" , hpx::util::format( "thread number %1% is out of range", num_thread)); return empty_mask; } // }}} mask_cref_type topology::get_core_affinity_mask( std::size_t num_thread , error_code& ec ) const { std::size_t num_pu = num_thread % num_of_pus_; if (num_pu < core_affinity_masks_.size()) { if (&ec != &throws) ec = make_success_code(); return core_affinity_masks_[num_pu]; } HPX_THROWS_IF(ec, bad_parameter , "hpx::threads::topology::get_core_affinity_mask" , hpx::util::format( "thread number %1% is out of range", num_thread)); return empty_mask; } mask_cref_type topology::get_thread_affinity_mask( std::size_t num_thread , error_code& ec ) const { // {{{ std::size_t num_pu = num_thread % num_of_pus_; if (num_pu < thread_affinity_masks_.size()) { if (&ec != &throws) ec = make_success_code(); return thread_affinity_masks_[num_pu]; } HPX_THROWS_IF(ec, bad_parameter , "hpx::threads::topology::get_thread_affinity_mask" , hpx::util::format( "thread number %1% is out of range", num_thread)); return empty_mask; } // }}} /////////////////////////////////////////////////////////////////////////// void topology::set_thread_affinity_mask( mask_cref_type mask , error_code& ec ) const { // {{{ #if !defined(__APPLE__) // setting thread affinities is not supported by OSX hwloc_cpuset_t cpuset = hwloc_bitmap_alloc(); int const pu_depth = hwloc_get_type_or_below_depth(topo, HWLOC_OBJ_PU); for (std::size_t i = 0; i != mask_size(mask); ++i) { if (test(mask, i)) { hwloc_obj_t const pu_obj = hwloc_get_obj_by_depth(topo, pu_depth, unsigned(i)); HPX_ASSERT(i == detail::get_index(pu_obj)); hwloc_bitmap_set(cpuset, static_cast<unsigned int>(pu_obj->os_index)); } } { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); if (hwloc_set_cpubind(topo, cpuset, HWLOC_CPUBIND_STRICT | HWLOC_CPUBIND_THREAD)) { // Strict binding not supported or failed, try weak binding. if (hwloc_set_cpubind(topo, cpuset, HWLOC_CPUBIND_THREAD)) { boost::scoped_ptr<char> buffer(new char [1024]); hwloc_bitmap_snprintf(buffer.get(), 1024, cpuset); hwloc_bitmap_free(cpuset); HPX_THROWS_IF(ec, kernel_error , "hpx::threads::topology::set_thread_affinity_mask" , hpx::util::format( "failed to set thread affinity mask (" HPX_CPU_MASK_PREFIX "%x) for cpuset %s", mask, buffer.get())); return; } } } #if defined(__linux) || defined(linux) || defined(__linux__) || defined(__FreeBSD__) sleep(0); // Allow the OS to pick up the change. #endif hwloc_bitmap_free(cpuset); #endif // __APPLE__ if (&ec != &throws) ec = make_success_code(); } // }}} /////////////////////////////////////////////////////////////////////////// mask_type topology::get_thread_affinity_mask_from_lva( naming::address_type lva , error_code& ec ) const { // {{{ if (&ec != &throws) ec = make_success_code(); hwloc_membind_policy_t policy = ::HWLOC_MEMBIND_DEFAULT; hwloc_nodeset_t nodeset = hwloc_bitmap_alloc(); { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); int ret = hwloc_get_area_membind_nodeset(topo, reinterpret_cast<void const*>(lva), 1, nodeset, &policy, 0); if (-1 != ret) { hwloc_cpuset_t cpuset = hwloc_bitmap_alloc(); hwloc_cpuset_from_nodeset(topo, cpuset, nodeset); lk.unlock(); hwloc_bitmap_free(nodeset); mask_type mask = mask_type(); resize(mask, get_number_of_pus()); int const pu_depth = hwloc_get_type_or_below_depth(topo, HWLOC_OBJ_PU); for (unsigned int i = 0; std::size_t(i) != num_of_pus_; ++i) { hwloc_obj_t const pu_obj = hwloc_get_obj_by_depth(topo, pu_depth, i); unsigned idx = static_cast<unsigned>(pu_obj->os_index); if (hwloc_bitmap_isset(cpuset, idx) != 0) set(mask, detail::get_index(pu_obj)); } hwloc_bitmap_free(cpuset); return mask; } } hwloc_bitmap_free(nodeset); return empty_mask; } // }}} std::size_t topology::init_node_number( std::size_t num_thread, hwloc_obj_type_t type ) { // {{{ if (std::size_t(-1) == num_thread) return std::size_t(-1); std::size_t num_pu = (num_thread + pu_offset) % num_of_pus_; { hwloc_obj_t obj; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_PU, static_cast<unsigned>(num_pu)); HPX_ASSERT(num_pu == detail::get_index(obj)); } while (obj) { if (hwloc_compare_types(obj->type, type) == 0) { return detail::get_index(obj); } obj = obj->parent; } } return 0; } // }}} void topology::extract_node_mask( hwloc_obj_t parent , mask_type& mask ) const { // {{{ hwloc_obj_t obj; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_next_child(topo, parent, nullptr); } while (obj) { if (hwloc_compare_types(HWLOC_OBJ_PU, obj->type) == 0) { do { set(mask, detail::get_index(obj)); //-V106 { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_next_child(topo, parent, obj); } } while (obj != nullptr && hwloc_compare_types(HWLOC_OBJ_PU, obj->type) == 0); return; } extract_node_mask(obj, mask); std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_next_child(topo, parent, obj); } } // }}} std::size_t topology::extract_node_count( hwloc_obj_t parent , hwloc_obj_type_t type , std::size_t count ) const { // {{{ hwloc_obj_t obj; if(parent == nullptr) return count; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_next_child(topo, parent, nullptr); } while (obj) { if (hwloc_compare_types(type, obj->type) == 0) { /* do { ++count; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_next_child(topo, parent, obj); } } while (obj != nullptr && hwloc_compare_types(type, obj->type) == 0); return count; */ ++count; } count = extract_node_count(obj, type, count); std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_next_child(topo, parent, obj); } return count; } // }}} std::size_t topology::get_number_of_sockets() const { int nobjs = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_SOCKET); if(0 > nobjs) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_number_of_sockets" , "hwloc_get_nbobjs_by_type failed"); return std::size_t(nobjs); } return std::size_t(nobjs); } std::size_t topology::get_number_of_numa_nodes() const { int nobjs = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_NODE); if(0 > nobjs) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_number_of_numa_nodes" , "hwloc_get_nbobjs_by_type failed"); return std::size_t(nobjs); } return std::size_t(nobjs); } std::size_t topology::get_number_of_cores() const { int nobjs = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_CORE); // If num_cores is smaller 0, we have an error if (0 > nobjs) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_number_of_cores" , "hwloc_get_nbobjs_by_type(HWLOC_OBJ_CORE) failed"); return std::size_t(nobjs); } else if (0 == nobjs) { // some platforms report zero cores but might still report the // number of PUs nobjs = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_PU); if (0 > nobjs) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_number_of_cores" , "hwloc_get_nbobjs_by_type(HWLOC_OBJ_PU) failed"); return std::size_t(nobjs); } } // the number of reported cores/pus should never be zero either to // avoid division by zero, we should always have at least one core if (0 == nobjs) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_number_of_cores" , "hwloc_get_nbobjs_by_type reports zero cores/pus"); return std::size_t(nobjs); } return std::size_t(nobjs); } std::size_t topology::get_number_of_socket_pus( std::size_t num_socket ) const { hwloc_obj_t socket_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); socket_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_SOCKET, static_cast<unsigned>(num_socket)); } if (socket_obj) { HPX_ASSERT(num_socket == detail::get_index(socket_obj)); std::size_t pu_count = 0; return extract_node_count(socket_obj, HWLOC_OBJ_PU, pu_count); } return num_of_pus_; } std::size_t topology::get_number_of_numa_node_pus( std::size_t numa_node ) const { hwloc_obj_t node_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); node_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_NODE, static_cast<unsigned>(numa_node)); } if (node_obj) { HPX_ASSERT(numa_node == detail::get_index(node_obj)); std::size_t pu_count = 0; node_obj = detail::adjust_node_obj(node_obj); return extract_node_count(node_obj, HWLOC_OBJ_PU, pu_count); } return num_of_pus_; } std::size_t topology::get_number_of_core_pus( std::size_t core ) const { hwloc_obj_t core_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); core_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_CORE, static_cast<unsigned>(core)); } if (core_obj) { HPX_ASSERT(core == detail::get_index(core_obj)); std::size_t pu_count = 0; return extract_node_count(core_obj, HWLOC_OBJ_PU, pu_count); } return num_of_pus_; } std::size_t topology::get_number_of_socket_cores( std::size_t num_socket ) const { hwloc_obj_t socket_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); socket_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_SOCKET, static_cast<unsigned>(num_socket)); } if (socket_obj) { HPX_ASSERT(num_socket == detail::get_index(socket_obj)); std::size_t pu_count = 0; return extract_node_count(socket_obj, HWLOC_OBJ_CORE, pu_count); } return get_number_of_cores(); } std::size_t topology::get_number_of_numa_node_cores( std::size_t numa_node ) const { hwloc_obj_t node_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); node_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_NODE, static_cast<unsigned>(numa_node)); } if (node_obj) { HPX_ASSERT(numa_node == detail::get_index(node_obj)); std::size_t pu_count = 0; node_obj = detail::adjust_node_obj(node_obj); return extract_node_count(node_obj, HWLOC_OBJ_CORE, pu_count); } return get_number_of_cores(); } hwloc_bitmap_ptr topology::cpuset_to_nodeset( mask_cref_type mask) const { hwloc_bitmap_t cpuset = mask_to_bitmap(mask, HWLOC_OBJ_PU); hwloc_bitmap_t nodeset = hwloc_bitmap_alloc(); hwloc_cpuset_to_nodeset_strict(topo, cpuset, nodeset); hwloc_bitmap_free(cpuset); return std::make_shared<hpx::threads::hpx_hwloc_bitmap_wrapper>(nodeset); } namespace detail { void print_info(std::ostream& os, hwloc_obj_t obj, char const* name, bool comma) { if (comma) os << ", "; os << name; if (obj->logical_index != ~0x0u) os << "L#" << obj->logical_index; if (obj->os_index != ~0x0u) os << "(P#" << obj->os_index << ")"; } void print_info(std::ostream& os, hwloc_obj_t obj, bool comma = false) { switch (obj->type) { case HWLOC_OBJ_PU: print_info(os, obj, "PU ", comma); break; case HWLOC_OBJ_CORE: print_info(os, obj, "Core ", comma); break; case HWLOC_OBJ_SOCKET: print_info(os, obj, "Socket ", comma); break; case HWLOC_OBJ_NODE: print_info(os, obj, "Node ", comma); break; default: break; } } } void topology::print_affinity_mask(std::ostream& os, std::size_t num_thread, mask_cref_type m, const std::string &pool_name) const { boost::io::ios_flags_saver ifs(os); bool first = true; for(std::size_t i = 0; i != num_of_pus_; ++i) { hwloc_obj_t obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_PU, unsigned(i)); if (!obj) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::print_affinity_mask" , "object not found"); return; } if(!test(m, detail::get_index(obj))) //-V106 continue; if (first) { first = false; os << std::setw(4) << num_thread << ": "; //-V112 //-V128 } else { os << " "; } detail::print_info(os, obj); while(obj->parent) { detail::print_info(os, obj->parent, true); obj = obj->parent; } os << ", on pool \"" << pool_name << "\""; os << std::endl; } } mask_type topology::init_machine_affinity_mask() const { // {{{ mask_type machine_affinity_mask = mask_type(); resize(machine_affinity_mask, get_number_of_pus()); hwloc_obj_t machine_obj; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); machine_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_MACHINE, 0); } if (machine_obj) { extract_node_mask(machine_obj, machine_affinity_mask); return machine_affinity_mask; } HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::init_machine_affinity_mask" , "failed to initialize machine affinity mask"); return empty_mask; } // }}} mask_type topology::init_socket_affinity_mask_from_socket( std::size_t num_socket ) const { // {{{ // If we have only one or no socket, the socket affinity mask // spans all processors if (std::size_t(-1) == num_socket) return machine_affinity_mask_; hwloc_obj_t socket_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); socket_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_SOCKET, static_cast<unsigned>(num_socket)); } if (socket_obj) { HPX_ASSERT(num_socket == detail::get_index(socket_obj)); mask_type socket_affinity_mask = mask_type(); resize(socket_affinity_mask, get_number_of_pus()); extract_node_mask(socket_obj, socket_affinity_mask); return socket_affinity_mask; } return machine_affinity_mask_; } // }}} mask_type topology::init_numa_node_affinity_mask_from_numa_node( std::size_t numa_node ) const { // {{{ // If we have only one or no NUMA domain, the NUMA affinity mask // spans all processors if (std::size_t(-1) == numa_node) { return machine_affinity_mask_; } hwloc_obj_t numa_node_obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); numa_node_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_NODE, static_cast<unsigned>(numa_node)); } if (numa_node_obj) { HPX_ASSERT(numa_node == detail::get_index(numa_node_obj)); mask_type node_affinity_mask = mask_type(); resize(node_affinity_mask, get_number_of_pus()); numa_node_obj = detail::adjust_node_obj(numa_node_obj); extract_node_mask(numa_node_obj, node_affinity_mask); return node_affinity_mask; } return machine_affinity_mask_; } // }}} mask_type topology::init_core_affinity_mask_from_core( std::size_t core, mask_cref_type default_mask ) const { // {{{ if (std::size_t(-1) == core) return default_mask; hwloc_obj_t core_obj = nullptr; std::size_t num_core = (core + core_offset) % get_number_of_cores(); { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); core_obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_CORE, static_cast<unsigned>(num_core)); } if (core_obj) { HPX_ASSERT(num_core == detail::get_index(core_obj)); mask_type core_affinity_mask = mask_type(); resize(core_affinity_mask, get_number_of_pus()); extract_node_mask(core_obj, core_affinity_mask); return core_affinity_mask; } return default_mask; } // }}} mask_type topology::init_thread_affinity_mask( std::size_t num_thread ) const { // {{{ if (std::size_t(-1) == num_thread) { return get_core_affinity_mask(num_thread); } std::size_t num_pu = (num_thread + pu_offset) % num_of_pus_; hwloc_obj_t obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_PU, static_cast<unsigned>(num_pu)); } if (!obj) { return get_core_affinity_mask(num_thread); } HPX_ASSERT(num_pu == detail::get_index(obj)); mask_type mask = mask_type(); resize(mask, get_number_of_pus()); set(mask, detail::get_index(obj)); //-V106 return mask; } // }}} mask_type topology::init_thread_affinity_mask( std::size_t num_core, std::size_t num_pu ) const { // {{{ hwloc_obj_t obj = nullptr; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); int num_cores = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_CORE); // If num_cores is smaller 0, we have an error, it should never be zero // either to avoid division by zero, we should always have at least one // core if (num_cores <= 0) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::init_thread_affinity_mask" , "hwloc_get_nbobjs_by_type failed"); return empty_mask; } num_core = (num_core + core_offset) % std::size_t(num_cores); obj = hwloc_get_obj_by_type(topo, HWLOC_OBJ_CORE, static_cast<unsigned>(num_core)); } if (!obj) return empty_mask;//get_core_affinity_mask(num_thread, false); HPX_ASSERT(num_core == detail::get_index(obj)); num_pu %= obj->arity; //-V101 //-V104 mask_type mask = mask_type(); resize(mask, get_number_of_pus()); set(mask, detail::get_index(obj->children[num_pu])); //-V106 return mask; } // }}} /////////////////////////////////////////////////////////////////////////// void topology::init_num_of_pus() { num_of_pus_ = 1; { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); int num_of_pus = hwloc_get_nbobjs_by_type(topo, HWLOC_OBJ_PU); if (num_of_pus > 0) { num_of_pus_ = static_cast<std::size_t>(num_of_pus); } } } std::size_t topology::get_number_of_pus() const { return num_of_pus_; } /////////////////////////////////////////////////////////////////////////// mask_type topology::get_cpubind_mask(error_code& ec) const { hwloc_cpuset_t cpuset = hwloc_bitmap_alloc(); mask_type mask = mask_type(); resize(mask, get_number_of_pus()); { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); if (hwloc_get_cpubind(topo, cpuset, HWLOC_CPUBIND_THREAD)) { hwloc_bitmap_free(cpuset); HPX_THROWS_IF(ec, kernel_error , "hpx::threads::topology::get_cpubind_mask" , "hwloc_get_cpubind failed"); return empty_mask; } int const pu_depth = hwloc_get_type_or_below_depth(topo, HWLOC_OBJ_PU); for (unsigned int i = 0; i != num_of_pus_; ++i) //-V104 { hwloc_obj_t const pu_obj = hwloc_get_obj_by_depth(topo, pu_depth, i); unsigned idx = static_cast<unsigned>(pu_obj->os_index); if (hwloc_bitmap_isset(cpuset, idx) != 0) set(mask, detail::get_index(pu_obj)); } } hwloc_bitmap_free(cpuset); if (&ec != &throws) ec = make_success_code(); return mask; } mask_type topology::get_cpubind_mask(compat::thread& handle, error_code& ec) const { hwloc_cpuset_t cpuset = hwloc_bitmap_alloc(); mask_type mask = mask_type(); resize(mask, get_number_of_pus()); { std::unique_lock<hpx::util::spinlock> lk(topo_mtx); #if defined(HPX_MINGW) if (hwloc_get_thread_cpubind(topo, pthread_gethandle(handle.native_handle()), cpuset, HWLOC_CPUBIND_THREAD)) #else if (hwloc_get_thread_cpubind(topo, handle.native_handle(), cpuset, HWLOC_CPUBIND_THREAD)) #endif { hwloc_bitmap_free(cpuset); HPX_THROWS_IF(ec, kernel_error , "hpx::threads::topology::get_cpubind_mask" , "hwloc_get_cpubind failed"); return empty_mask; } int const pu_depth = hwloc_get_type_or_below_depth(topo, HWLOC_OBJ_PU); for (unsigned int i = 0; i != num_of_pus_; ++i) //-V104 { hwloc_obj_t const pu_obj = hwloc_get_obj_by_depth(topo, pu_depth, i); unsigned idx = static_cast<unsigned>(pu_obj->os_index); if (hwloc_bitmap_isset(cpuset, idx) != 0) set(mask, detail::get_index(pu_obj)); } } hwloc_bitmap_free(cpuset); if (&ec != &throws) ec = make_success_code(); return mask; } /////////////////////////////////////////////////////////////////////////// /// This is equivalent to malloc(), except that it tries to allocate /// page-aligned memory from the OS. void* topology::allocate(std::size_t len) const { return hwloc_alloc(topo, len); } /////////////////////////////////////////////////////////////////////////// /// Allocate some memory on NUMA memory nodes specified by nodeset /// as specified by the hwloc hwloc_alloc_membind_nodeset call void* topology::allocate_membind(std::size_t len, hwloc_bitmap_ptr bitmap, hpx_hwloc_membind_policy policy, int flags) const { return hwloc_alloc_membind_nodeset(topo, len, bitmap->get_bmp(), (hwloc_membind_policy_t)(policy), flags); } bool topology::set_area_membind_nodeset( const void *addr, std::size_t len, void *nodeset) const { hwloc_membind_policy_t policy = ::HWLOC_MEMBIND_BIND; hwloc_nodeset_t ns = reinterpret_cast<hwloc_nodeset_t>(nodeset); int ret = hwloc_set_area_membind_nodeset(topo, addr, len, ns, policy, 0); if (ret<0) { std::string msg = std::strerror(errno); if (errno == ENOSYS) msg = "the action is not supported"; if (errno == EXDEV) msg = "the binding cannot be enforced"; HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::set_area_membind_nodeset" , "hwloc_set_area_membind_nodeset failed : " + msg); return false; } return true; } util::thread_specific_ptr<hpx_hwloc_bitmap_wrapper, topology::tls_tag> topology::bitmap_storage_; threads::mask_type topology::get_area_membind_nodeset( const void *addr, std::size_t len) const { hpx_hwloc_bitmap_wrapper *nodeset = topology::bitmap_storage_.get(); if (nullptr == nodeset) { hwloc_bitmap_t nodeset_ = hwloc_bitmap_alloc(); topology::bitmap_storage_.reset(new hpx_hwloc_bitmap_wrapper(nodeset_)); nodeset = topology::bitmap_storage_.get(); } // hwloc_membind_policy_t policy; hwloc_nodeset_t ns = reinterpret_cast<hwloc_nodeset_t>(nodeset->get_bmp()); if (hwloc_get_area_membind_nodeset(topo, addr, len, ns, &policy, 0)==-1) { HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_area_membind_nodeset" , "hwloc_get_area_membind_nodeset failed"); return -1; std::cout << "error in " ; } return bitmap_to_mask(ns, HWLOC_OBJ_NUMANODE); } int topology::get_numa_domain(const void *addr) const { #if HWLOC_API_VERSION >= 0x00010b03 hpx_hwloc_bitmap_wrapper *nodeset = topology::bitmap_storage_.get(); if (nullptr == nodeset) { hwloc_bitmap_t nodeset_ = hwloc_bitmap_alloc(); topology::bitmap_storage_.reset(new hpx_hwloc_bitmap_wrapper(nodeset_)); nodeset = topology::bitmap_storage_.get(); } // hwloc_nodeset_t ns = reinterpret_cast<hwloc_nodeset_t>(nodeset->get_bmp()); int ret = hwloc_get_area_memlocation(topo, addr, 1, ns, HWLOC_MEMBIND_BYNODESET); if (ret<0) { std::string msg(strerror(errno)); HPX_THROW_EXCEPTION(kernel_error , "hpx::threads::topology::get_numa_domain" , "hwloc_get_area_memlocation failed " + msg); return -1; } threads::mask_type mask = bitmap_to_mask(ns, HWLOC_OBJ_NUMANODE); return threads::find_first(mask); #else return 0; #endif } /// Free memory that was previously allocated by allocate void topology::deallocate(void* addr, std::size_t len) const { hwloc_free(topo, addr, len); } /////////////////////////////////////////////////////////////////////////// hwloc_bitmap_t topology::mask_to_bitmap(mask_cref_type mask, hwloc_obj_type_t htype) const { hwloc_bitmap_t bitmap = hwloc_bitmap_alloc(); hwloc_bitmap_zero(bitmap); // int const depth = hwloc_get_type_or_below_depth(topo, htype); for (std::size_t i = 0; i != mask_size(mask); ++i) { if (test(mask, i)) { hwloc_obj_t const hw_obj = hwloc_get_obj_by_depth(topo, depth, unsigned(i)); HPX_ASSERT(i == detail::get_index(hw_obj)); hwloc_bitmap_set(bitmap, static_cast<unsigned int>(hw_obj->os_index)); } } return bitmap; } /////////////////////////////////////////////////////////////////////////// mask_type topology::bitmap_to_mask(hwloc_bitmap_t bitmap, hwloc_obj_type_t htype) const { mask_type mask = mask_type(); std::size_t num = hwloc_get_nbobjs_by_type(topo, htype); // int const pu_depth = hwloc_get_type_or_below_depth(topo, htype); for (unsigned int i=0; std::size_t(i)!=num; ++i) //-V104 { hwloc_obj_t const pu_obj = hwloc_get_obj_by_depth(topo, pu_depth, i); unsigned idx = static_cast<unsigned>(pu_obj->os_index); if (hwloc_bitmap_isset(bitmap, idx) != 0) set(mask, detail::get_index(pu_obj)); } return mask; } /////////////////////////////////////////////////////////////////////////// void topology::print_mask_vector(std::ostream& os, std::vector<mask_type> const& v) const { std::size_t s = v.size(); if (s == 0) { os << "(empty)\n"; return; } for (std::size_t i = 0; i != s; i++) { os << std::hex << HPX_CPU_MASK_PREFIX << v[i] << "\n"; } os << "\n"; } void topology::print_vector( std::ostream& os, std::vector<std::size_t> const& v) const { std::size_t s = v.size(); if (s == 0) { os << "(empty)\n"; return; } os << v[0]; for (std::size_t i = 1; i != s; i++) { os << ", " << std::dec << v[i]; } os << "\n"; } void topology::print_hwloc(std::ostream& os) const { os << "[HWLOC topology info] number of ...\n" << std::dec << "number of sockets : " << get_number_of_sockets() << "\n" << "number of numa nodes : " << get_number_of_numa_nodes() << "\n" << "number of cores : " << get_number_of_cores() << "\n" << "number of PUs : " << get_number_of_pus() << "\n" << "hardware concurrency : " << hpx::threads::hardware_concurrency() << "\n" << std::endl; //! -------------------------------------- topology (affinity masks) os << "[HWLOC topology info] affinity masks :\n" << "machine : \n" << std::hex << HPX_CPU_MASK_PREFIX << machine_affinity_mask_ << "\n"; os << "socket : \n"; print_mask_vector(os, socket_affinity_masks_); os << "numa node : \n"; print_mask_vector(os, numa_node_affinity_masks_); os << "core : \n"; print_mask_vector(os, core_affinity_masks_); os << "PUs (/threads) : \n"; print_mask_vector(os, thread_affinity_masks_); //! -------------------------------------- topology (numbers) os << "[HWLOC topology info] resource numbers :\n"; os << "socket : \n"; print_vector(os, socket_numbers_); os << "numa node : \n"; print_vector(os, numa_node_numbers_); os << "core : \n"; print_vector(os, core_numbers_); //os << "PUs (/threads) : \n"; //print_vector(os, pu_numbers_); } topology const& get_topology() { hpx::runtime* rt = hpx::get_runtime_ptr(); if (rt == nullptr) { HPX_THROW_EXCEPTION(invalid_status, "hpx::threads::get_topology", "the hpx runtime system has not been initialized yet"); } return rt->get_topology(); } /////////////////////////////////////////////////////////////////////////// struct hardware_concurrency_tag {}; struct hw_concurrency { hw_concurrency() #if defined(__ANDROID__) && defined(ANDROID) : num_of_cores_(::android_getCpuCount()) #else : num_of_cores_(detail::hwloc_hardware_concurrency()) #endif { if (num_of_cores_ == 0) num_of_cores_ = 1; } std::size_t num_of_cores_; }; std::size_t hardware_concurrency() { util::static_<hw_concurrency, hardware_concurrency_tag> hwc; return hwc.get().num_of_cores_; } }}
{ "alphanum_fraction": 0.5428243935, "author": null, "avg_line_length": 32.0482993197, "converted": null, "ext": "cpp", "file": null, "hexsha": "4f6c540155e5f21f22a67428977b3c0c7838c89c", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "d1db0def60687a662e4e25a909550f08eaf1a18a", "max_forks_repo_licenses": [ "BSL-1.0" ], "max_forks_repo_name": "victor-ludorum/hpx", "max_forks_repo_path": "src/runtime/threads/topology.cpp", "max_issues_count": 1, "max_issues_repo_head_hexsha": "a1621a0cfa58a884b03bc8557d4f5ad896297f14", "max_issues_repo_issues_event_max_datetime": "2018-04-20T14:17:33.000Z", "max_issues_repo_issues_event_min_datetime": "2018-04-20T14:17:33.000Z", "max_issues_repo_licenses": [ "BSL-1.0" ], "max_issues_repo_name": "ShmuelLevine/hpx", "max_issues_repo_path": "src/runtime/threads/topology.cpp", "max_line_length": 88, "max_stars_count": null, "max_stars_repo_head_hexsha": "a1621a0cfa58a884b03bc8557d4f5ad896297f14", "max_stars_repo_licenses": [ "BSL-1.0" ], "max_stars_repo_name": "ShmuelLevine/hpx", "max_stars_repo_path": "src/runtime/threads/topology.cpp", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 11159, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 47111 }
# coding: utf-8 # In[7]: from __future__ import absolute_import from __future__ import print_function import keras from keras.preprocessing.image import ImageDataGenerator from keras.layers import Dense, Dropout, Activation, Flatten from keras.layers import Conv2D, MaxPooling2D from keras import applications from keras.preprocessing.image import ImageDataGenerator from keras import optimizers from keras.models import Sequential, Model from keras.layers import Dropout, Flatten, Dense, GlobalAveragePooling2D from keras import backend as k from keras.callbacks import ModelCheckpoint, LearningRateScheduler, TensorBoard, EarlyStopping import numpy as np import pandas as pd import urllib from sklearn.cross_validation import train_test_split import glob import os import pickle from trainer.environment import create_trainer_environment # the trainer environment contains useful information about env = create_trainer_environment() print('creating SageMaker trainer environment:\n%s' % str(env)) ## LOAD the data from train and test channels npX_keras = pickle.load(open(os.path.join(env.channel_dirs['train'], 'npX_keras.pkl'), "rb")) oh_npY = pickle.load(open(os.path.join(env.channel_dirs['train'], 'oh_npY.pkl'), "rb")) #Lest split the data into train and validation train_X, validation_X, train_y, validation_y = train_test_split(npX_keras, oh_npY, test_size=0.2, random_state=1001) batch_size = 32 epochs = 1 model = applications.VGG19(weights = "imagenet", include_top=False, input_shape = (128, 128, 3)) # Freeze the layers which you don't want to train. Here I am freezing the first 5 layers. for layer in model.layers[:5]: layer.trainable = False #Adding custom Layers x = model.output x = Flatten()(x) x = Dense(1024, activation="relu")(x) x = Dropout(0.5)(x) x = Dense(1024, activation="relu")(x) predictions = Dense(12, activation="softmax")(x) # creating the final model model_final = Model(input = model.input, output = predictions) # In[57]: # Initiate the train and test generators with data Augumentation img_width, img_height = 128, 128 train_datagen = ImageDataGenerator( rescale = 1./255, horizontal_flip = True, fill_mode = "nearest", zoom_range = 0.3, width_shift_range = 0.3, height_shift_range=0.3, rotation_range=30) test_datagen = ImageDataGenerator( rescale = 1./255, horizontal_flip = True, fill_mode = "nearest", zoom_range = 0.3, width_shift_range = 0.3, height_shift_range=0.3, rotation_range=30) train_generator = train_datagen.flow(train_X, train_y, batch_size=batch_size) validation_generator = test_datagen.flow(validation_X, validation_y) # Save the model according to the conditions #checkpoint = ModelCheckpoint("vgg16_1.h5", monitor='val_acc', verbose=1, # save_best_only=True, save_weights_only=False, mode='auto', period=1) #early = EarlyStopping(monitor='val_acc', min_delta=0, patience=10, verbose=0, mode='auto') # ## PREP for AWS Sagemaker - start.py # In[30]: MODEL_NAME = 'seedling_model.h5' # getting the hyperparameters batch_size = env.hyperparameters.get('batch_size', object_type=int) learning_rate = env.hyperparameters.get('learning_rate', default=.0001, object_type=float) EPOCHS = env.hyperparameters.get('epochs', default=10, object_type=int) # TRAIN Model n_train_samples = train_X.shape[0] n_validation_samples = validation_X.shape[0] *1.0 # compile the model #model_final.compile(loss="categorical_crossentropy", # optimizer=optimizers.Adam(lr=learning_rate), metrics=["accuracy"]) model_final.compile(loss="categorical_crossentropy", optimizer=optimizers.SGD(lr=learning_rate, momentum=0.9), metrics=["accuracy"]) model_final.fit_generator( train_generator, samples_per_epoch=n_train_samples/batch_size, epochs=EPOCHS, validation_data=validation_generator, validation_steps=n_validation_samples/batch_size) # Save model and weights model_path = os.path.join(env.model_dir, MODEL_NAME) model_final.save(model_path) print('Saved trained model at %s ' % model_path) # Score trained model. scores = model_final.evaluate_generator(validation_generator, n_validation_samples/batch_size, verbose=1) print('Test loss:', scores[0]) print('Test accuracy:', scores[1])
{ "alphanum_fraction": 0.7049956934, "author": null, "avg_line_length": 32.9361702128, "converted": null, "ext": "py", "file": null, "hexsha": "759774f3d3d6a3fc8d8140381570bd61e664f568", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2020-03-04T18:16:57.000Z", "max_forks_repo_forks_event_min_datetime": "2020-03-04T18:16:57.000Z", "max_forks_repo_head_hexsha": "6ce1ad1acc80420b15ef043b6f18b9a6bca8e737", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "ThePrecious/ml_projects", "max_forks_repo_path": "3_Sagemaker_Seedling/trainer/start.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "6ce1ad1acc80420b15ef043b6f18b9a6bca8e737", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "ThePrecious/ml_projects", "max_issues_repo_path": "3_Sagemaker_Seedling/trainer/start.py", "max_line_length": 116, "max_stars_count": 3, "max_stars_repo_head_hexsha": "6ce1ad1acc80420b15ef043b6f18b9a6bca8e737", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "ThePrecious/ml_projects", "max_stars_repo_path": "3_Sagemaker_Seedling/trainer/start.py", "max_stars_repo_stars_event_max_datetime": "2019-06-22T17:48:38.000Z", "max_stars_repo_stars_event_min_datetime": "2018-10-10T20:32:11.000Z", "num_tokens": 1054, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4644 }
#!/usr/bin/env python # coding: utf-8 import os import numpy as np import pandas as pd from matplotlib import pyplot as plt import seaborn as sb sb.set(style="white") sb.set(color_codes=True) sb.set_context('paper') os.chdir('/Users/pauline/Documents/Python') df = pd.read_csv("Tab-Bathy.csv") # define variables and plotting g = sb.clustermap(df, cmap="BuPu") rotation = 45 for i, ax in enumerate(g.fig.axes): # getting all axes of the fig object ax.set_xticklabels(ax.get_xticklabels(), rotation = rotation ) g.fig.suptitle('Mariana Trench: Clustermap of the bathymetric observations. \nDataset: 25 cross-section profiles, 518 observations in each') plt.subplots_adjust(bottom=0.20, top=0.90, right=0.90, left=0.10 ) # printing and saving plt.savefig('plot_Clust3.png', dpi=300) plt.show()
{ "alphanum_fraction": 0.6756756757, "author": null, "avg_line_length": 28.6451612903, "converted": null, "ext": "py", "file": null, "hexsha": "d5c887b723027fd64248cc2f672296905ecffd1e", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "b305f4b344f631d1c459d4f62bbd9587cc51c91f", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "paulinelemenkova/Python-script-011-Clustermap", "max_forks_repo_path": "Script-011-Clustermap-3.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "b305f4b344f631d1c459d4f62bbd9587cc51c91f", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "paulinelemenkova/Python-script-011-Clustermap", "max_issues_repo_path": "Script-011-Clustermap-3.py", "max_line_length": 140, "max_stars_count": null, "max_stars_repo_head_hexsha": "b305f4b344f631d1c459d4f62bbd9587cc51c91f", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "paulinelemenkova/Python-script-011-Clustermap", "max_stars_repo_path": "Script-011-Clustermap-3.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 228, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 888 }
module numz !************************************* integer, parameter:: b8 = selected_real_kind(14) ! basic real types integer, parameter:: b4 = selected_real_kind(4) ! integer, parameter:: i8 = selected_int_kind(14) integer, parameter:: out1 = 11 ! output file number for writes integer, parameter:: out2 = 20 integer, parameter:: in1 = 13 integer, parameter:: in2 = 14 ! real(b8), parameter :: pi = 3.141592653589793239_b8 end module
{ "alphanum_fraction": 0.6316916488, "author": null, "avg_line_length": 35.9230769231, "converted": null, "ext": "f90", "file": null, "hexsha": "446922b4774ae04c8bf460371ec6e22897976767", "include": null, "lang": "FORTRAN", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "04c162ec890a1c9ba83498b275fbdc81a4704062", "max_forks_repo_licenses": [ "Unlicense" ], "max_forks_repo_name": "timkphd/examples", "max_forks_repo_path": "darwin19/numz.f90", "max_issues_count": 1, "max_issues_repo_head_hexsha": "04c162ec890a1c9ba83498b275fbdc81a4704062", "max_issues_repo_issues_event_max_datetime": "2022-02-09T01:59:47.000Z", "max_issues_repo_issues_event_min_datetime": "2022-02-09T01:59:47.000Z", "max_issues_repo_licenses": [ "Unlicense" ], "max_issues_repo_name": "timkphd/examples", "max_issues_repo_path": "darwin19/numz.f90", "max_line_length": 71, "max_stars_count": 5, "max_stars_repo_head_hexsha": "04c162ec890a1c9ba83498b275fbdc81a4704062", "max_stars_repo_licenses": [ "Unlicense" ], "max_stars_repo_name": "timkphd/examples", "max_stars_repo_path": "darwin19/numz.f90", "max_stars_repo_stars_event_max_datetime": "2022-01-24T19:09:47.000Z", "max_stars_repo_stars_event_min_datetime": "2020-11-01T00:29:22.000Z", "num_tokens": 130, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 467 }
""" Set of functions for measuring the spectral properties of galaxies. """ from __future__ import (absolute_import, division, print_function, unicode_literals) from ..plot import Plot from astropy.stats import sigma_clip from astropy.constants import c from astropy.units import km, s, erg, cm, angstrom from astropy.modeling.models import Gaussian1D, GaussianAbsorption1D from astropy.modeling.fitting import LevMarLSQFitter from math import fabs import matplotlib.pyplot as pl import numpy as np import random __all__ = ['line_measurements'] c_kms = c.to(km / s).value transition_wavelengths = {'OII': 3728.484, 'Hd': 4102.890, 'Hg': 4341.680, 'Hb': 4862.680, 'OIIIB': 4960.295, 'OIIIR': 5008.240, 'NIIB': 6549.840, 'Ha': 6564.610, 'NIIR': 6585.230, 'SIIB': 6718.320, 'SIIR': 6732.710} # Vacuum wavelengths o2_doublet = (3727.09, 3729.88) o2 = (7580, 7680) oh = (8600, 8700) sigma_max = 10 bbox = dict(facecolor='w', edgecolor='none') TwoGaussians = (Gaussian1D + Gaussian1D).rename('TwoGaussians') ThreeGaussians = (Gaussian1D + Gaussian1D + Gaussian1D).rename('ThreeGaussians') def tie_sigma(model): stddev = model.stddev_0 return stddev def tie_sii(model): amplitude = model.amplitude_0 / 1.4 return amplitude def tie_nii(model): amplitude = model.amplitude_0 / 2.95 return amplitude def tie_oiii(model): amplitude = model.amplitude_0 / 2.98 return amplitude def clean_spectrum(wavelength, flux, error): inf = np.isinf(error) zero = flux == 0 nan = np.isnan(flux) cond = ~inf & ~zero & ~nan return wavelength[cond], flux[cond], error[cond], cond def monte_carlo_error(wavelength, flux, error, continuum, fitter, init, absorption=False, n=100): perturbed_g = [] for i in range(0, n): if absorption: perturb = np.array([random.gauss(0, 1) * error[i] / continuum[i] for i in range(len(error))]) flux_perturbed = (flux / continuum) + perturb else: perturb = np.array([random.gauss(0, 1) * error[i] for i in range(len(error))]) flux_perturbed = flux - continuum + perturb perturbed_g.append(fitter(init, wavelength, flux_perturbed)) params = {} for i, name in enumerate(init.param_names): params[name] = [] for g in perturbed_g: for i, name in enumerate(g.param_names): params[name].append(g.parameters[i]) errors = {} for key in params.keys(): errors[key] = np.std(np.array(params[key])) return errors def line_measurements(name, spec1d, z, sky=None, spec2d=None, resolution=200, yposition=None, sky_threshold=None, fit_o2_doublet=False, plot=False, show_plot=False, plot_directory=None, sky_in_counts=False): """ Measures emission line fluxes and equivalent widths from a galaxy spectrum with redshift z. Parameters ---------- name : str Object ID. spec1d : `igmtools.data.Spectrum1D` 1D spectrum. z : float Galaxy redshift. sky : array 1D sky spectrum. spec2d : `igmtools.data.Spectrum2D` 2D spectrum. R : int, optional Spectral resolution (default = 200). yposition : tuple, optional y-coordinates of the object on the 2D spectrum (edge 1, centre, edge 2). sky_threshold : float, optional Sky counts/flux value above which flag 3 is raised (useful for eliminating zero-order contamination and regions of bad sky subtraction). fit_o2_doublet : bool, optional Option to fit two Gaussian components to the OII line (default = False). plot : bool, optional Option to plot continuum estimates across bands where a measurement is executed, and the Gaussian fit, if performed (default = False). show_plot : bool, optional Option to show each plot in an interactive window (default = False). plot_directory : str, optional If specified, plots will be saved in this directory as PNG files. sky_in_counts : bool, optional Set to True if the sky spectrum is in counts rather than flux units. Returns ------- measurements : `astropy.table.Table` The line measurements. Notes ----- Measurements are made using the following line indicies: OII : (3655.0, 3705.0, 3708.5, 3748.5, 3750.0, 3800.0) Hd : (4030.0, 4080.0, 4082.0, 4122.0, 4125.0, 4170.0) Hg : (4230.0, 4270.0, 4321.5, 4361.5, 4365.0, 4400.0) Hb : (4785.0, 4820.0, 4842.5, 4882.5, 5030.0, 5100.0) OIII : (4785.0, 4820.0, 4988.0, 5028.0, 5030.0, 5100.0) Ha : (6460.0, 6520.0, 6544.5, 6584.5, 6610.0, 6670.0), SII : (6640.0, 6700.0, 6713.0, 6753.0, 6760.0, 6810.0) These line indicies are optimised for spectra with R ~ 200, and have measurement windows 20 Angstroms wide. We assume the maximum intrinsic line width to be that of a Gaussian with a standard deviation of 10 Angstroms in the rest frame. Convolved with the instrument line spread function, this corresponds to measurement windows of width close to the maximum expected standard deviation of the Gaussian line profile. For specified instrument resolutions different to the default value of 200, measurement windows are scaled to preserve this feature. Lines with integrated fluxes measured at greater than 3 sigma significance are fitted with Gaussians. The measurements are then taken from the Gaussian fitting parameters and their errors computed from a Monte-Carlo type estimation. The maximum allowed Gaussian standard deviation corresponds to 10 Angstroms in the intrinsic line profile, and the minimum to 0.5 times that of that instrumental line spread function. If Hdelta is absorption dominated at a greater than 3 sigma level, a Gaussian absorption profile is fitted. This is motivated by the idea that Hdelta may be used as a proxy for the Balmer absorption correction. Equivalent widths are positive for emission lines, and negative for absorption lines. Warning flags are defined as follows: 0 : No warnings. 1 : Measurement may be affected by the OH forest between 8600 and 8700 Angstrom. 2 : Line was fit with the maximum/minimum allowed Gaussian standard deviation. 3 : Line coincides with region above the specified sky threshold. 4 : Line may be affected by O2 telluric absorption (7580 - 7680 Angstrom). 5 : Bad continuum reduced chi squared (> 10). 6 : No spectral coverage, or human verification failed. No measurement is recorded for flag 6 - all values are set to -99.0. """ plot = True if show_plot else plot bands = {'OII': [3655.0, 3705.0, 3708.5, 3748.5, 3750.0, 3800.0], 'Hd': [4030.0, 4080.0, 4082.0, 4122.0, 4125.0, 4170.0], 'Hg': [4230.0, 4270.0, 4321.5, 4361.5, 4365.0, 4400.0], 'Hb': [4785.0, 4820.0, 4842.5, 4882.5, 5030.0, 5100.0], 'OIII': [4785.0, 4820.0, 4988.0, 5028.0, 5030.0, 5100.0], 'Ha': [6460.0, 6520.0, 6544.5, 6584.5, 6610.0, 6670.0], 'SII': [6640.0, 6700.0, 6713.0, 6753.0, 6760.0, 6810.0]} # Modify measurement windows if appropriate: if resolution != 200: sigma_max200 = 18.8 # Max Gaussian sigma for R = 200 (approx) dlambda = 7500 / resolution sigma_lsf = dlambda / 2.35482 sigma_max_convolved = np.sqrt(sigma_lsf ** 2 + sigma_max ** 2) scale_factor = sigma_max_convolved / sigma_max200 for key in bands.keys(): window = bands[key][3] - bands[key][2] window0 = window * scale_factor bands[key][1] += (window - window0) / 2 bands[key][2] += (window - window0) / 2 bands[key][3] -= (window - window0) / 2 bands[key][4] -= (window - window0) / 2 # Initialise dictionaries: (line_flux, continuum_flux, eqw, sn, continuum_params, line_params, flags) = {}, {}, {}, {}, {}, {}, {} # 1D spectrum arrays: wavelength = spec1d.wavelength.value flux = spec1d.flux.value error = spec1d.flux.uncertainty.value # Clean the spectrum: wavelength, flux, error, cond = clean_spectrum(wavelength, flux, error) if sky is not None: sky = sky[cond] # Do measurements: for key in bands.keys(): # Initialise dictionary for continuum parameters: continuum_params[key] = {} # Line groupings: if (key == 'OII') and fit_o2_doublet: lines = ['OIIR', 'OIIB'] rest_wavelengths = o2_doublet elif key == 'OIII': lines = ['OIIIR', 'OIIIB'] rest_wavelengths = [transition_wavelengths['OIIIR'], transition_wavelengths['OIIIB']] elif key == 'Ha': lines = ['NIIR', 'Ha', 'NIIB'] rest_wavelengths = [transition_wavelengths['NIIR'], transition_wavelengths['Ha'], transition_wavelengths['NIIB']] elif key == 'SII': lines = ['SIIR', 'SIIB'] rest_wavelengths = [transition_wavelengths['SIIR'], transition_wavelengths['SIIB']] else: lines = [key] rest_wavelengths = [transition_wavelengths[key]] # Initialise dictionaries for line parameters: for line in lines: line_params[line] = {} # Observed wavelengths of the lines: observed_wavelengths = [item * (1 + z) for item in rest_wavelengths] # Fitting/measurement regions: co_blue = ((wavelength >= bands[key][0] * (1 + z)) & (wavelength < bands[key][1] * (1 + z))) co_red = ((wavelength >= bands[key][4] * (1 + z)) & (wavelength < bands[key][5] * (1 + z))) co_region = co_red | co_blue line_region = ((wavelength >= bands[key][2] * (1 + z)) & (wavelength <= bands[key][3] * (1 + z))) # Extended region around the measurement. Used for excluding # measurements affected by zero orders: centre = ((bands[key][2] * (1 + z)) + (bands[key][3] * (1 + z))) / 2 centre_band = ((wavelength >= centre - 100) & (wavelength <= centre + 100)) # The full fitting region: region = ((wavelength >= bands[key][0] * (1 + z)) & (wavelength < bands[key][5] * (1 + z))) # Masks to identify regions potentially affected by 7600A O2 telluric # absorption: o2_blue = ((wavelength[co_blue] >= o2[0]) & (wavelength[co_blue] <= o2[1])) o2_red = (wavelength[co_red] >= o2[0]) & (wavelength[co_red] <= o2[1]) o2_line = ((wavelength[line_region] >= o2[0]) & (wavelength[line_region] <= o2[1])) # Masks to identify regions potentially affected by the OH forest: oh_blue = ((wavelength[co_blue] >= oh[0]) & (wavelength[co_blue] <= oh[1])) oh_red = (wavelength[co_red] >= oh[0]) & (wavelength[co_red] <= oh[1]) oh_line = ((wavelength[line_region] >= oh[0]) & (wavelength[line_region] <= oh[1])) # Assume the measurement will be good at first: flags[key] = 0 # Check that we have spectral coverage: if ((np.sum(co_blue) < 5) | (np.sum(co_red) < 5) | (flux[co_blue] == 0).all() | (flux[co_red] == 0).all()): # If no coverage, mark all measurements as -99.0 and assign flag # 6, then go to next iteration of the loop: for line in lines: line_flux[line] = (-99.0, -99.0) continuum_flux[line] = (-99.0, -99.0) eqw[line] = (-99.0, -99.0) line_params[line]['amplitude'] = -99.0 line_params[line]['mean'] = -99.0 line_params[line]['stddev'] = -99.0 continuum_params[key]['gradient'] = -99.0 continuum_params[key]['intercept'] = -99.0 continuum_params[key]['chi2norm'] = -99.0 sn[key] = -99.0 flags[key] = 6 continue # See if we're affected by 7600A O2 telluric absorption: if ((np.sum(o2_blue) > 0) | (np.sum(o2_red) > 0) | (np.sum(o2_line) > 0)): flags[key] = 4 # See if we're affected by OH forest: if ((np.sum(oh_blue) > 0) | (np.sum(oh_red) > 0) | (np.sum(oh_line) > 0)): flags[key] = 1 # Assign sky threshold flag if a value is specified and it exceeds # this: if sky_threshold is not None: if any(sky0 > sky_threshold for sky0 in sky[centre_band]): flags[key] = 3 # Sigma clip the continuum, to ensure it's not affected by nearby # absorption features: filtered_blue = sigma_clip(flux[co_blue], 1.5) filtered_red = sigma_clip(flux[co_red], 1.5) # Take the mean value of the sigma clipped continuum either side of the # line: co_level1 = np.mean(filtered_blue) co_level1_error = np.std(filtered_blue) co_level2 = np.mean(filtered_red) co_level2_error = np.std(filtered_red) # Linearly interpolate between these values: continuum = ((co_level2 - co_level1) / (np.mean(wavelength[co_red]) - np.mean(wavelength[co_blue])) * (wavelength - np.mean(wavelength[co_blue])) + co_level1) continuum_error = np.sqrt( co_level1_error ** 2 + co_level2_error ** 2) / 2 # Continuum gradient: gradient = ((continuum[1] - continuum[0]) / (wavelength[1] - wavelength[0])) continuum_params[key]['gradient'] = gradient # Continuum intercept: intercept = continuum[0] - gradient * wavelength[0] continuum_params[key]['intercept'] = intercept # Flag if normalised continuum chi squared > 10: cont_chi2norm = (np.sum( (flux[co_region] - continuum[co_region]) ** 2 / error[co_region] ** 2) / (len(flux[co_region]) - 3)) if cont_chi2norm > 10: flags[key] = 5 continuum_params[key]['chi2norm'] = cont_chi2norm # Estimate integrated line flux and equivalent width (observed, # not rest frame): dl = np.mean(wavelength[line_region][1:] - wavelength[line_region][:-1]) n = np.sum(line_region) line_flux_value = dl * np.sum( flux[line_region] - continuum[line_region]) line_flux_error = dl * np.sqrt( np.sum(error[line_region] ** 2) + n * continuum_error ** 2) eqw_value = dl * np.sum( flux[line_region] / continuum[line_region]) - n eqw_error = dl * np.sqrt( np.sum(error[line_region] ** 2 / continuum[line_region] ** 2) + n * continuum_error ** 2 * np.sum(flux[line_region] ** 2 / continuum[line_region] ** 4)) # Continuum flux at the line centre: ind = np.abs(wavelength - observed_wavelengths[0]).argmin() centre_flux = continuum[ind] centre_flux_error = error[ind] # Estimate signal-to-noise ratio around the line: sn_blue = (filtered_blue[~filtered_blue.mask] / error[co_blue][~filtered_blue.mask]) sn_red = (filtered_red[~filtered_red.mask] / error[co_red][~filtered_red.mask]) sn_value = np.average(np.concatenate([sn_blue, sn_red])) # Calculate minimum and maximum allowed Gaussian standard # deviations: dlambda = rest_wavelengths[0] / resolution sigma_lsf = dlambda / 2.35482 min_stddev = sigma_lsf / 2 max_stddev = np.sqrt(sigma_lsf ** 2 + sigma_max ** 2) * (1 + z) # Fit Gaussian component(s) if the integrated line flux is # positive and has greater than 3 sigma significance: if (line_flux_value > 0) & ((line_flux_value / line_flux_error) > 3): amplitude = np.max(flux[line_region] - continuum[line_region]) if ((key in ('Hg', 'Hd')) | ((key == 'OII') and not fit_o2_doublet)): # One component Gaussian fit for Hg, Hd, OII: # ------------------------------------------- mean = observed_wavelengths[0] g_init = Gaussian1D(amplitude, mean, min_stddev) g_init.amplitude.min = 0.0 g_init.mean.min = mean - (1000 * mean / c_kms) g_init.mean.max = mean + (1000 * mean / c_kms) g_init.stddev.min = min_stddev g_init.stddev.max = max_stddev # ------------------------------------------- elif (key == 'OII') and fit_o2_doublet: # Optional two component Gaussian fit for OII: # -------------------------------------------- mean_0 = observed_wavelengths[0] mean_1 = observed_wavelengths[1] tied_params = {'stddev_1': tie_sigma} g_init = TwoGaussians( amplitude, mean_0, min_stddev, amplitude, mean_1, min_stddev, tied=tied_params) g_init.amplitude_0.min = 0.0 g_init.amplitude_1.min = 0.0 g_init.mean_0.min = mean_0 - (1000 * mean_0 / c_kms) g_init.mean_0.max = mean_0 + (1000 * mean_0 / c_kms) g_init.mean_1.min = mean_1 - (1000 * mean_1 / c_kms) g_init.mean_1.max = mean_1 + (1000 * mean_1 / c_kms) g_init.stddev_0.min = min_stddev g_init.stddev_0.max = max_stddev g_init.stddev_1.min = min_stddev g_init.stddev_1.max = max_stddev # -------------------------------------------- elif key == 'SII': # Two component Gaussian fit for SII: # ----------------------------------- mean_0 = observed_wavelengths[0] mean_1 = observed_wavelengths[1] tied_params = {'amplitude_1': tie_sii, 'stddev_1': tie_sigma} g_init = TwoGaussians( amplitude, mean_0, min_stddev, amplitude, mean_1, min_stddev, tied=tied_params) g_init.amplitude_0.min = 0.0 g_init.amplitude_1.min = 0.0 g_init.mean_0.min = mean_0 - (1000 * mean_0 / c_kms) g_init.mean_0.max = mean_0 + (1000 * mean_0 / c_kms) g_init.mean_1.min = mean_1 - (1000 * mean_1 / c_kms) g_init.mean_1.max = mean_1 + (1000 * mean_1 / c_kms) g_init.stddev_0.min = min_stddev g_init.stddev_0.max = max_stddev g_init.stddev_1.min = min_stddev g_init.stddev_1.max = max_stddev # ----------------------------------- elif key in ('Hb', 'OIII'): # Three component fit over Hb/OIII region: # ---------------------------------------- mean_0 = observed_wavelengths[0] if key == 'Hb': mean_1 = transition_wavelengths['OIIIB'] * (1 + z) mean_2 = transition_wavelengths['OIIIR'] * (1 + z) tied_params = {'stddev_1': tie_sigma, 'stddev_2': tie_sigma} else: mean_1 = transition_wavelengths['OIIIB'] * (1 + z) mean_2 = transition_wavelengths['Hb'] * (1 + z) tied_params = {'amplitude_1': tie_oiii, 'stddev_1': tie_sigma, 'stddev_2': tie_sigma} g_init = ThreeGaussians( amplitude, mean_0, min_stddev, amplitude, mean_1, min_stddev, amplitude, mean_2, min_stddev, tied=tied_params) g_init.amplitude_0.min = 0.0 g_init.amplitude_1.min = 0.0 g_init.amplitude_2.min = 0.0 g_init.mean_0.min = mean_0 - (1000 * mean_0 / c_kms) g_init.mean_0.max = mean_0 + (1000 * mean_0 / c_kms) g_init.mean_1.min = mean_1 - (1000 * mean_1 / c_kms) g_init.mean_1.max = mean_1 + (1000 * mean_1 / c_kms) g_init.mean_2.min = mean_2 - (1000 * mean_2 / c_kms) g_init.mean_2.max = mean_2 + (1000 * mean_2 / c_kms) g_init.stddev_0.min = min_stddev g_init.stddev_0.max = max_stddev g_init.stddev_1.min = min_stddev g_init.stddev_1.max = max_stddev g_init.stddev_2.min = min_stddev g_init.stddev_2.max = max_stddev # ---------------------------------------- else: # Try one and three component fit over Ha/NII region: # --------------------------------------------------- mean_1 = observed_wavelengths[1] g_init = Gaussian1D(amplitude, mean_1, min_stddev) g_init.amplitude.min = 0.0 g_init.mean.min = mean_1 - (1000 * mean_1 / c_kms) g_init.mean.max = mean_1 + (1000 * mean_1 / c_kms) g_init.stddev.min = min_stddev g_init.stddev.max = max_stddev mean_0 = observed_wavelengths[0] mean_2 = observed_wavelengths[2] tied_params = {'amplitude_2': tie_nii, 'stddev_1': tie_sigma, 'stddev_2': tie_sigma} g_init2 = ThreeGaussians( amplitude, mean_0, min_stddev, amplitude, mean_1, min_stddev, amplitude, mean_2, min_stddev, tied=tied_params) g_init2.amplitude_0.min = 0.0 g_init2.amplitude_1.min = 0.0 g_init2.amplitude_2.min = 0.0 g_init2.mean_0.min = mean_0 - (1000 * mean_0 / c_kms) g_init2.mean_0.max = mean_0 + (1000 * mean_0 / c_kms) g_init2.mean_1.min = mean_1 - (1000 * mean_1 / c_kms) g_init2.mean_1.max = mean_1 + (1000 * mean_1 / c_kms) g_init2.mean_2.min = mean_2 - (1000 * mean_2 / c_kms) g_init2.mean_2.max = mean_2 + (1000 * mean_2 / c_kms) g_init2.stddev_0.min = min_stddev g_init2.stddev_0.max = max_stddev g_init2.stddev_1.min = min_stddev g_init2.stddev_1.max = max_stddev g_init2.stddev_2.min = min_stddev g_init2.stddev_2.max = max_stddev # --------------------------------------------------- # Do the fitting: fit_g = LevMarLSQFitter() g = fit_g(g_init, wavelength[region], flux[region] - continuum[region]) # Chi2 on the fit: line_chi2 = np.sum( (flux[region] - continuum[region] - g(wavelength)[region]) ** 2 / error[region] ** 2) # Monte carlo error estimation: g_errors = monte_carlo_error( wavelength[region], flux[region], error[region], continuum[region], fit_g, g_init) ha_3comp = False # Compare chi squared values for the two Ha/NII fits and adopt # the one that has the minimum chi squared: if key == 'Ha': # Three component fit of Ha/NII region: fit_g2 = LevMarLSQFitter() g2 = fit_g2(g_init2, wavelength[region], flux[region] - continuum[region]) # Monte carlo error estimation: g2_errors = monte_carlo_error( wavelength[region], flux[region], error[region], continuum[region], fit_g2, g_init2) # Chi2 on the fit: line_chi2_2 = np.sum( (flux[region] - continuum[region] - g2(wavelength[region])) ** 2 / (g2(wavelength[region]) + continuum[region])) # Compare chi2: if line_chi2 > line_chi2_2: g = g2 g_errors = g2_errors ha_3comp = True # Get lists of best-fit Gaussian parameters: if ((key in ('Hg', 'Hd')) | ((key == 'OII') and not fit_o2_doublet)): amplitudes = [g.amplitude.value] amplitude_errors = [g_errors['amplitude']] means = [g.mean.value] stddevs = [g.stddev.value] stddev_errors = [g_errors['stddev']] elif ((key == 'OII') and fit_o2_doublet) | (key == 'SII'): amplitudes = [g.amplitude_0.value, g.amplitude_1.value] amplitude_errors = [g_errors['amplitude_0'], g_errors['amplitude_1']] means = [g.mean_0.value, g.mean_1.value] stddevs = [g.stddev_0.value, g.stddev_1.value] stddev_errors = [g_errors['stddev_0'], g_errors['stddev_1']] elif ((key == 'Ha') and ha_3comp) | (key in ('Hb', 'OIII')): amplitudes = [g.amplitude_0.value, g.amplitude_1.value, g.amplitude_2.value] amplitude_errors = [g_errors['amplitude_0'], g_errors['amplitude_1'], g_errors['amplitude_2']] means = [g.mean_0.value, g.mean_1.value, g.mean_2.value] stddevs = [g.stddev_0.value, g.stddev_1.value, g.mean_2.value] stddev_errors = [g_errors['stddev_0'], g_errors['stddev_1'], g_errors['stddev_2']] else: amplitudes = [g.amplitude.value, -99.0, -99.0] amplitude_errors = [g_errors['amplitude'], -99.0, -99.0] means = [g.mean.value, -99.0, -99.0] stddevs = [g.stddev.value, -99.0, -99.0] stddev_errors = [g_errors['stddev'], -99.0, -99.0] # Log these line by line: for i, line in enumerate(lines): # Log the line fitting parameters: line_params[line]['amplitude'] = amplitudes[i] line_params[line]['mean'] = means[i] line_params[line]['stddev'] = stddevs[i] # Only adopt the measurements if the fitted amplitude is # non-zero, otherwise, measurements from direct integration # of pixels are retained: if amplitudes[i] != 0: # Integrated line flux: line_flux_value = (amplitudes[i] * stddevs[i] * np.sqrt(2 * np.pi)) # Error on the integrated line flux: line_flux_error = line_flux_value * np.sqrt( (amplitude_errors[i] / amplitudes[i]) ** 2 + (stddev_errors[i] / stddevs[i]) ** 2) # Re-evaluate the continuum flux at the line centre: ind = np.abs(wavelength - means[i]).argmin() centre_flux = continuum[ind] centre_flux_error = error[ind] # Equivalent width: eqw_value = line_flux_value / centre_flux # Error on the equivalent width: eqw_error = eqw_value * np.sqrt( (line_flux_error / line_flux_value) ** 2 + (centre_flux_error / centre_flux) ** 2) # Log the line flux, continuum flux and equivalent width: line_flux[line] = (line_flux_value, line_flux_error) continuum_flux[line] = (centre_flux, centre_flux_error) eqw[line] = (eqw_value, eqw_error) fit = True fit_hd = False # Fit single Gaussian absorption component to Hd if the integrated # line flux is negative and has greater than 3 sigma significance: elif ((key == 'Hd') & (line_flux_value < 0) & ((line_flux_value / line_flux_error) < -3)): amplitude = np.max(1 - flux[line_region] / continuum[line_region]) mean = transition_wavelengths[key] * (1 + z) dm = 1000 * mean / c_kms g_init = GaussianAbsorption1D(amplitude, mean, min_stddev) g_init.mean.min = mean - dm g_init.mean.max = mean + dm g_init.stddev.min = min_stddev g_init.stddev.max = max_stddev # Do the fitting: fit_g = LevMarLSQFitter() g = fit_g(g_init, wavelength[region], flux[region] / continuum[region]) # Monte carlo error estimation: g_errors = monte_carlo_error( wavelength[region], flux[region], error[region], continuum[region], fit_g, g_init, absorption=True) # Equivalent width: eqw_value = (-g.amplitude.value * g.stddev.value * np.sqrt(2 * np.pi) / (1 + z)) # Error on the equivalent width: eqw_error = fabs(eqw_value) * np.sqrt( (g_errors['amplitude'] / g.amplitude.value) ** 2 + (g_errors['stddev'] / g.stddev.value) ** 2) for line in lines: # Log the line fitting parameters: line_params[line]['amplitude'] = g.amplitude.value line_params[line]['mean'] = g.mean.value line_params[line]['stddev'] = g.stddev.value # Log the line flux, continuum flux and equivalent width: line_flux[line] = (line_flux_value, line_flux_error) continuum_flux[line] = (centre_flux, centre_flux_error) eqw[line] = (eqw_value, eqw_error) fit = False fit_hd = True # Otherwise we won't do any line fitting: else: for line in lines: # Set all line fitting parameters to -99: line_params[line]['amplitude'] = -99.0 line_params[line]['mean'] = -99.0 line_params[line]['stddev'] = -99.0 # Log the line flux, continuum flux and equivalent width: line_flux[line] = (line_flux_value, line_flux_error) continuum_flux[line] = (centre_flux, centre_flux_error) eqw[line] = (eqw_value, eqw_error) fit = False fit_hd = False sn[key] = sn_value # Make plots if that option is turned on: if plot: if sky is not None and spec2d is not None: n = 3 p = Plot(n, 1, n, aspect=1, width=5.9, fontsize=12) elif ((sky is not None and spec2d is None) | (spec2d is not None and sky is None)): n = 2 p = Plot(n, 1, n, aspect=0.8, width=5.9, fontsize=12) else: n = 1 p = Plot(n, 1, n, aspect=0.6, width=5.9, fontsize=12) centre = (bands[key][0] * (1 + z) + bands[key][5] * (1 + z)) / 2 cond = (wavelength > centre - 250) & (wavelength < centre + 250) if spec2d is not None: cond2 = ((spec2d.wavelength.value > centre - 250) & (spec2d.wavelength.value < centre + 250)) # 2D spectrum plot: if spec2d is not None: n = 3 if n == 3 else 2 # 2D spectrum parameters for plotting: i = min(spec2d.data.shape[0] // 2, 3) v1 = np.percentile(spec2d.data[i:-1, :].ravel(), 90) wdelt = spec2d.wavelength.value[1] - spec2d.wavelength.value[0] yvals = np.arange(spec2d.data.shape[0]) * wdelt p.axes[n - n].pcolormesh( spec2d.wavelength.value[cond2], yvals, spec2d.data[:, cond2], vmin=-v1 / 5, vmax=2 * v1, cmap=pl.cm.hot) if yposition is not None: p.axes[n - n].axhline( wdelt * yposition[0], ls='--', lw=2, color='LawnGreen') p.axes[n - n].axhline( wdelt * yposition[2], ls='--', lw=2, color='LawnGreen') # 1D spectrum plot: if (n == 3) | ((n == 2) & (sky is not None and spec2d is None)): n = 2 else: n = 1 p.axes[n - n].plot( wavelength[cond], flux[cond] / 1e-16, drawstyle='steps-mid', color='k') p.axes[n - n].plot( wavelength[cond], error[cond] / 1e-16, drawstyle='steps-mid', color='r') p.axes[n - n].plot( wavelength[region], continuum[region] / 1e-16, lw=3, color='RoyalBlue') if fit: p.axes[n - n].plot( wavelength[region], (g(wavelength[region]) + continuum[region]) / 1e-16, color='m', lw=2) if fit_hd: p.axes[n - n].plot( wavelength[region], (g(wavelength[region]) * continuum[region]) / 1e-16, color='m', lw=2) p.axes[n - n].axvspan( bands[key][2] * (1 + z), bands[key][3] * (1 + z), facecolor='g', edgecolor='none', alpha=0.5) p.axes[n - n].annotate( key, xy=(0.05, 0.8), xycoords='axes fraction', horizontalalignment='left', fontsize=12, bbox=bbox, color='k') p.axes[n - n].annotate( name, xy=(0.95, 0.8), xycoords='axes fraction', horizontalalignment='right', fontsize=12, bbox=bbox, color='k') # Sky spectrum plot: if sky is not None: if sky_in_counts: p.axes[n - 1].plot( wavelength[cond], sky[cond], drawstyle='steps-mid', color='MidnightBlue') else: p.axes[n - 1].plot( wavelength[cond], sky[cond] / 1e-16, drawstyle='steps-mid', color='MidnightBlue') p.axes[n - 1].annotate( 'sky', xy=(0.05, 0.8), xycoords='axes fraction', horizontalalignment='left', fontsize=12, bbox=bbox, color='k') # Axis limits and labels, tidy up and display: region_min = np.min(error[region]) region_max = np.max(flux[region]) sn_spec = np.median(flux / error) for i in range(0, n): p.axes[i].set_xlim(wavelength[cond][0], wavelength[cond][-1]) p.axes[n - 2].set_ylim( (region_min - 0.2 * sn_spec * region_min) / 1e-16, (region_max + 0.8 * region_max) / 1e-16) xlabel = 'Wavelength ($\AA$)' if spec1d.flux.unit == erg / s / cm ** 2 / angstrom: ylabel = 'Flux ($10^{-16}$ erg s$^{-1}$ cm$^{-2}$ $\AA$)' else: ylabel = 'Flux ({0})'.format( spec1d.flux.unit.to_string(format='latex')) p.tidy(shared_axes=True) p.labels(xlabel, ylabel) if plot_directory is not None: p.savefig('{0}/{1}_{2}.png'.format(plot_directory, name, key)) if show_plot: p.display() return (line_flux, continuum_flux, eqw, sn, continuum_params, line_params, flags)
{ "alphanum_fraction": 0.5232984649, "author": null, "avg_line_length": 38.4185803758, "converted": null, "ext": "py", "file": null, "hexsha": "e127479671e43975de018607f981b85443308ffd", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2019-11-19T04:45:38.000Z", "max_forks_repo_forks_event_min_datetime": "2019-11-19T04:45:38.000Z", "max_forks_repo_head_hexsha": "6e14973fd1e69d5e7bd7c40f93ffe11e2cd41990", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "cwfinn/igmtools", "max_forks_repo_path": "igmtools/modeling/galaxy.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "6e14973fd1e69d5e7bd7c40f93ffe11e2cd41990", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "cwfinn/igmtools", "max_issues_repo_path": "igmtools/modeling/galaxy.py", "max_line_length": 79, "max_stars_count": null, "max_stars_repo_head_hexsha": "6e14973fd1e69d5e7bd7c40f93ffe11e2cd41990", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "cwfinn/igmtools", "max_stars_repo_path": "igmtools/modeling/galaxy.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 9417, "path": null, "reason": "import numpy,from astropy", "repo": null, "save_path": null, "sha": null, "size": 36805 }
#include <turbodbc/field_translators/timestamp_translator.h> #include <boost/variant/get.hpp> #include <sql.h> namespace turbodbc { namespace field_translators { timestamp_translator::timestamp_translator() = default; timestamp_translator::~timestamp_translator() = default; field timestamp_translator::do_make_field(char const * data_pointer) const { auto const ts = reinterpret_cast<SQL_TIMESTAMP_STRUCT const *>(data_pointer); // map SQL nanosecond precision to posix_time microsecond precision int64_t const adjusted_fraction = ts->fraction / 1000; return {boost::posix_time::ptime{ {static_cast<short unsigned int>(ts->year), ts->month, ts->day}, {ts->hour, ts->minute, ts->second, adjusted_fraction} } }; } } }
{ "alphanum_fraction": 0.741130092, "author": null, "avg_line_length": 28.1851851852, "converted": null, "ext": "cpp", "file": null, "hexsha": "fb8a4e01d2f1faa4950c1586e9f2d6e0f72beadc", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 83, "max_forks_repo_forks_event_max_datetime": "2022-03-26T02:32:26.000Z", "max_forks_repo_forks_event_min_datetime": "2016-06-15T13:55:44.000Z", "max_forks_repo_head_hexsha": "80a29a7edfbdabf12410af01c0c0ae74bfc3aab4", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "arikfr/turbodbc", "max_forks_repo_path": "cpp/turbodbc/Library/src/field_translators/timestamp_translator.cpp", "max_issues_count": 325, "max_issues_repo_head_hexsha": "80a29a7edfbdabf12410af01c0c0ae74bfc3aab4", "max_issues_repo_issues_event_max_datetime": "2022-03-21T23:58:42.000Z", "max_issues_repo_issues_event_min_datetime": "2016-04-08T11:54:41.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "arikfr/turbodbc", "max_issues_repo_path": "cpp/turbodbc/Library/src/field_translators/timestamp_translator.cpp", "max_line_length": 78, "max_stars_count": 537, "max_stars_repo_head_hexsha": "80a29a7edfbdabf12410af01c0c0ae74bfc3aab4", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "arikfr/turbodbc", "max_stars_repo_path": "cpp/turbodbc/Library/src/field_translators/timestamp_translator.cpp", "max_stars_repo_stars_event_max_datetime": "2022-03-29T04:43:17.000Z", "max_stars_repo_stars_event_min_datetime": "2016-03-18T21:46:05.000Z", "num_tokens": 183, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 761 }
Base.@kwdef struct FieldInfo name::String ltype::Union{DataType,String,Symbol} rtype::Union{DataType,String,Symbol} = rtype get_returns_reference::Bool = true end struct TypeInfo name::String fields::Vector{FieldInfo} end
{ "alphanum_fraction": 0.7217741935, "author": null, "avg_line_length": 19.0769230769, "converted": null, "ext": "jl", "file": null, "hexsha": "bfc9a32bc789087a238cc615b203ff17549c939e", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "53b5fa074338a5eea096d6736453c5692faf7368", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "colinxs/Emporos.jl", "max_forks_repo_path": "archive/CodeGen/types.jl", "max_issues_count": null, "max_issues_repo_head_hexsha": "53b5fa074338a5eea096d6736453c5692faf7368", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "colinxs/Emporos.jl", "max_issues_repo_path": "archive/CodeGen/types.jl", "max_line_length": 48, "max_stars_count": null, "max_stars_repo_head_hexsha": "53b5fa074338a5eea096d6736453c5692faf7368", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "colinxs/Emporos.jl", "max_stars_repo_path": "archive/CodeGen/types.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 68, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 248 }
""" Kernel PCA object """ struct kPCA{T <: Real} X::Matrix{Union{T, Missing}} # original data κ::Kernel{T} # kernel function (from MLKernels) μ::Vector{T} # row/col means of the kernel matrix μ2::T # mean of the kernel matrix λ::Vector{T} # eigenvalues in feature space α::Matrix{T} # eigenvectors in feature space function kPCA( X :: Matrix{Union{T, Missing}}, κ :: Kernel{T}, μ :: Vector{T}, μ2 :: T, λ :: Vector{T}, α :: Matrix{T} ) where T <: Real @assert length(λ) == size(α, 2) @assert length(μ) == size(X, 2) new{T}(X, κ, μ, μ2, λ, α) end end indim(M::kPCA) = size(M.X, 1) outdim(M::kPCA) = length(M.λ) projection(M::kPCA) = M.V ./ sqrt.(M.λ) principalvars(M::kPCA) = M.λ """ fit a kernel PCA. κ :: Kernel{T} the kernel function, see MLKernels.jl for documentation. X :: Array{Union{T, Missing}} the data to embed, with observations ins columns. """ function fit( :: Type{kPCA}, κ :: Kernel{T}, X :: Matrix{Union{T, Missing}}, ncomp = 2 ) where T K = kernelmatrix(Val(:col), κ, X, true) K2, μ, μ2 = missing_to_mean(K) K2 .= K2 .- μ .- μ' .+ μ2 e = K2 |> Hermitian |> eigen e_ev_perm = sortperm(e.values, rev = true)[1:ncomp] λ = e.values[e_ev_perm]::Vector{T} α = e.vectors[:, e_ev_perm]::Matrix{T} return kPCA(X, κ, μ, μ2, λ, α) end """Calculate transformation to kernel space""" # function transform(M::kPCA{T}, x::AbstractMatrix{mT}) where mT <: Union{T, Missing} where T <: Real function transform(M::kPCA{T}, x::AbstractMatrix{mT}) where mT where T ##### There is some typeinference bug that leads to illegal instructions and ##### reaches the unreachable... K = kernelmatrix(Val(:col), M.κ, M.X, x) # ##### instead of `kernelmatrix`: # # K = allocate_basematrix(Val(:col), M.X, x) # K = Array{T}(undef, size(M.X, 2), size(x, 2)) # f = basefunction(M.κ) # # basematrix!(Val(:col), K, f, M.X, x) # ##### instead of `basematrix!`: # n, m = checkdimensions(Val(:col), K, M.X, x) # @inbounds for j = 1:m # yj = subvector(Val(:col), x, j) # for i = 1:n # xi = subvector(Val(:col), M.X, i) # ##### this causes slightly less allocations but far worse performance at least it doesn't crash the julia session # # # unsafe_base_evaluate # # s, nn = base_initiate(f, T), 0 # # # s, nn = base_initiate(f, promote_type(eltype(x), eltype(y))), 0 # # for i in eachindex(xi, yj) # # s, nn = base_aggregate(f, s, nn, xi[i], yj[i]) # # end # # K[i, j] = nn > 0 ? base_return(f, s / nn) : missing # # # end unsafe_base_evaluate # K[i,j] = unsafe_base_evaluate(f, xi, yj) # end # end # ##### end instead of `basematrix!` # kappamatrix!(M.κ, K) # ##### end instead of `kernelmatrix` K .= K .- M.μ .- mean(K, dims = 1) .+ M.μ2 return (M.α' ./ sqrt.(M.λ)) * K end transform(M::kPCA) = sqrt.(M.λ) .* M.α' function Base.show(io::IO, ::MIME"text/plain", M::kPCA{T}) where T println(io, "kPCA{", T, "}(κ: ", M.κ, ", dims: ", indim(M), "─→", outdim(M), ")") end # """Kernel PCA type""" # struct KernelPCA{T<:Real} # X::AbstractMatrix{T} # fitted data or precomputed kernel # ker::Union{Nothing, Function} # kernel function # center::KernelCenter # kernel center # λ::AbstractVector{T} # eigenvalues in feature space # α::AbstractMatrix{T} # eigenvectors in feature space # inv::AbstractMatrix{T} # inverse transform coefficients # end # struct KernelCenter{T<:Real} # means::AbstractVector{T} # total::T # end # import MultivariateStats: KernelPCA, KernelCenter, transform! # import Base.convert # """Center kernel matrix.""" # function transform!(C::KernelCenter{T}, K::AbstractMatrix{T}) where {T<:Real} # r, c = size(K) # tot = C.total # means = mean(K, dims=1) # K .= K .- C.means .- means .+ tot # return K # end # TODO: does not work # function convert(::Type{KernelPCA{T}}, M::kPCA) where T # KernelPCA(missing_to_mean_slice(M.X, dims = 2), # (x, y) -> kappa(M.κ, unsafe_base_evaluate(basefunction(M.κ), x, y)), # KernelCenter(M.μ, M.μ2), # M.λ, # M.α, # zeros(T, 0, 0)) # end
{ "alphanum_fraction": 0.5645454545, "author": null, "avg_line_length": 30.5555555556, "converted": null, "ext": "jl", "file": null, "hexsha": "37caff6d0cddd649c7ed5fdb50bd4c496000b1d2", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "fe6b98d48be7be86d2f7f000cfa8407d37f43199", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "gdkrmr/MLKernelsMissing.jl", "max_forks_repo_path": "src/kpca.jl", "max_issues_count": null, "max_issues_repo_head_hexsha": "fe6b98d48be7be86d2f7f000cfa8407d37f43199", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "gdkrmr/MLKernelsMissing.jl", "max_issues_repo_path": "src/kpca.jl", "max_line_length": 115, "max_stars_count": null, "max_stars_repo_head_hexsha": "fe6b98d48be7be86d2f7f000cfa8407d37f43199", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "gdkrmr/MLKernelsMissing.jl", "max_stars_repo_path": "src/kpca.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1416, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 4400 }
from keras.models import load_model from load_face_dataset import load_dataset, IMAGE_SIZE, resize_image import numpy as np from fr_utils import img_to_encoding from sklearn.model_selection import cross_val_score, ShuffleSplit, KFold from sklearn.neighbors import KNeighborsClassifier from sklearn.externals import joblib import matplotlib.pyplot as plt import os os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE" from keras.utils import CustomObjectScope import tensorflow as tf with CustomObjectScope({'tf': tf}): facenet = load_model('./model/nn4.small2.v1.h5') class Dataset: def __init__(self, path_name): self.X_train = None self.y_train = None self.path_name = path_name def load(self, img_rows = IMAGE_SIZE, img_cols = IMAGE_SIZE, img_channels = 3, model = facenet): images, labels = load_dataset(self.path_name) X_embedding = img_to_encoding(images, model) print('X_train shape', X_embedding.shape) print('y_train shape', labels.shape) print(X_embedding.shape[0], 'train samples') self.X_train = X_embedding self.y_train = labels class Knn_Model: def __init__(self): self.model = None def cross_val_and_build_model(self, dataset): k_range = range(1,31) k_scores = [] print("k vs accuracy:") for k in k_range: knn = KNeighborsClassifier(n_neighbors = k) cv = KFold(n_splits = 10, shuffle = True, random_state = 0) score = cross_val_score(knn, dataset.X_train, dataset.y_train, cv = 10, scoring = 'accuracy').mean() k_scores.append(score) print(k, ":", score) plt.plot(k_range, k_scores) plt.title("KNN") #fontsize = 24) plt.xlabel('Value of K for KNN')#, fontsize = 14) plt.ylabel('Cross-Validated Accuracy')#, fontsize = 14) plt.tick_params(axis='both')#, labelsize = 14) plt.show() n_neighbors_max = np.argmax(k_scores) + 1 print("The best k is: ", n_neighbors_max) print("The accuracy is: ", k_scores[n_neighbors_max - 1], "When n_neighbor is: ", n_neighbors_max) self.model = KNeighborsClassifier(n_neighbors = n_neighbors_max) def train(self, dataset): self.model.fit(dataset.X_train, dataset.y_train) def save_model(self, file_path): #save model joblib.dump(self.model, file_path) def load_model(self, file_path): self.model = joblib.load(file_path) def predict(self, image): image = resize_image(image) image_embedding = img_to_encoding(np.array([image]), facenet) label = self.model.predict(image_embedding) return label[0] if __name__ == "__main__": dataset = Dataset('dataset') dataset.load() model = Knn_Model() model.cross_val_and_build_model(dataset) model.train(dataset) model.save_model('model/knn_classifier.model')
{ "alphanum_fraction": 0.6368642695, "author": null, "avg_line_length": 29.1226415094, "converted": null, "ext": "py", "file": null, "hexsha": "70ef1ecf78fd65ee39c652583df96736a0de9d0d", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "ecb1fed88ab5bbb969446266f3326ce12ac086ee", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "donghu123/face_recognition_using_opencv_keras_scikit-learn-master", "max_forks_repo_path": "face_knn_classifier.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "ecb1fed88ab5bbb969446266f3326ce12ac086ee", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "donghu123/face_recognition_using_opencv_keras_scikit-learn-master", "max_issues_repo_path": "face_knn_classifier.py", "max_line_length": 112, "max_stars_count": null, "max_stars_repo_head_hexsha": "ecb1fed88ab5bbb969446266f3326ce12ac086ee", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "donghu123/face_recognition_using_opencv_keras_scikit-learn-master", "max_stars_repo_path": "face_knn_classifier.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 713, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 3087 }
import os import numpy as np from pwtools import io from pwtools.test import tools from .testenv import testdir rand = np.random.rand def test_h5(): try: import h5py dct1 = \ {'/a': 'abcgs', '/b/c/x1': 3, '/b/c/x2': rand(2,3), } # writing a dct w/o leading slash will always be read back in *with* # leading slash dct2 = \ {'a': 'abciqo4iki', 'b/c/x1': 3, 'b/c/x2': rand(2,3), } for idx,dct in enumerate([dct1, dct2]): h5fn = os.path.join(testdir, 'test_%i.h5' %idx) io.write_h5(h5fn, dct) read_dct = io.read_h5(h5fn) for kk in list(read_dct.keys()): assert kk.startswith('/') for kk in list(dct.keys()): key = '/'+kk if not kk.startswith('/') else kk tools.assert_all_types_equal(dct[kk], read_dct[key]) # write mode='a' is default, test appending h5fn = os.path.join(testdir, 'test_append.h5') io.write_h5(h5fn, {'/a': 1.0}) read_dct = io.read_h5(h5fn) assert list(read_dct.keys()) == ['/a'] assert read_dct['/a'] == 1.0 # append '/b', using {'/a': 1.0, '/b': 2.0} would be an error since /a # already exists, use mode='w' then, but this overwrites all! io.write_h5(h5fn, {'/b': 2.0}) read_dct2 = io.read_h5(h5fn) # sort(...): sort possible [/b, /a] -> [/a, /b] assert np.sort(np.array(list(read_dct2.keys()))).tolist() == ['/a', '/b'] assert read_dct2['/a'] == 1.0 assert read_dct2['/b'] == 2.0 # overwrite io.write_h5(h5fn, {'/b': 22.0, '/c': 33.0}, mode='w') read_dct3 = io.read_h5(h5fn) assert np.sort(np.array(list(read_dct3.keys()))).tolist() == ['/b', '/c'] except ImportError: tools.skip("skipping test_h5, no h5py importable")
{ "alphanum_fraction": 0.5114854518, "author": null, "avg_line_length": 36.2777777778, "converted": null, "ext": "py", "file": null, "hexsha": "19e08c50ec4f9d13ec3ea8cb5a1b03fd9284a0ae", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 9, "max_forks_repo_forks_event_max_datetime": "2022-02-20T01:52:03.000Z", "max_forks_repo_forks_event_min_datetime": "2017-06-01T12:57:57.000Z", "max_forks_repo_head_hexsha": "cee068d1c7984d85e94ace243f86de350d3a1dba", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "elcorto/pwtools", "max_forks_repo_path": "pwtools/test/test_h5.py", "max_issues_count": 4, "max_issues_repo_head_hexsha": "cee068d1c7984d85e94ace243f86de350d3a1dba", "max_issues_repo_issues_event_max_datetime": "2021-10-30T21:12:53.000Z", "max_issues_repo_issues_event_min_datetime": "2019-02-27T09:14:17.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "elcorto/pwtools", "max_issues_repo_path": "pwtools/test/test_h5.py", "max_line_length": 81, "max_stars_count": 41, "max_stars_repo_head_hexsha": "cee068d1c7984d85e94ace243f86de350d3a1dba", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "elcorto/pwtools", "max_stars_repo_path": "pwtools/test/test_h5.py", "max_stars_repo_stars_event_max_datetime": "2022-02-24T16:08:47.000Z", "max_stars_repo_stars_event_min_datetime": "2016-06-25T13:17:57.000Z", "num_tokens": 604, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1959 }
import time import math import numpy as np from itertools import product from collections import deque, defaultdict, namedtuple #TODO #4 make 2nd part of the task def load_data(): with open("AoC2020/aoc20-20-data.txt", "r") as f: data = f.read() print("Loaded lines:", len(data.splitlines())) with open("AoC2020/aoc20-20-testdata.txt", "r") as f: testdata = f.read() print("Loaded lines:", len(testdata.splitlines())) return data, testdata Shape = namedtuple('Shape',['N','E','S','W','arr']) def make_shape(arr): # borders = [arr[:,0], arr[0,:], arr[:,9], arr[9,:]] borders = [arr[0,:], arr[:,9], arr[9,:], arr[:,0]] num_borders = [] for b in borders: num = int(''.join(map(str,b)),2) num_borders.append(num) sh = Shape(*num_borders,np.copy(arr)) return sh def get_shapes(array): shapes = [] arr = array for _ in range(2): for _ in range(4): shapes.append(make_shape(arr)) arr = np.rot90(arr) arr = np.fliplr(arr) return tuple(shapes) Tile = namedtuple('Tile',['id','array','shapes','edges']) def parse_data(p_data): tiles = [] segments = data.split('\n\n') for segment in segments: header, pattern = segment.split(':\n') tile_id=int(header[5:]) bin_list=list(pattern.replace('\n','').replace('#','1').replace('.','0')) arr = np.array(bin_list,dtype=np.int8).reshape(10,10) shapes=get_shapes(arr) edges = None tiles.append(Tile(tile_id, arr,shapes, edges)) return tiles def fit(mosaic, shape, px,py): if not 0<=px<len(mosaic) or not 0<=py<len(mosaic): return False if px>0 and mosaic[px-1][py]: if mosaic[px-1][py][1].E != shape.W: return False if px<len(mosaic)-1 and mosaic[px+1][py]: if mosaic[px+1][py][1].W != shape.E: return False if py>0 and mosaic[px][py-1]: if mosaic[px][py-1][1].N != shape.S: return False if py<len(mosaic)-1 and mosaic[px][py+1]: if mosaic[px][py+1][1].S != shape.N: return False return True def dfs_helper(mosaic, graf, px, py, pstack, tile, edge_value): result = False matching_tiles = graf.get(edge_value) for item in matching_tiles.values(): if item != tile: result = dfs(mosaic, graf, item, px, py, pstack) return result def dfs(mosaic, graf, tile, px, py, pstack): if not 0<=px<len(mosaic) or not 0<=py<len(mosaic) or mosaic[px][py]: return False if tile not in pstack: pstack.append(tile) for shape in tile.shapes: if fit(mosaic, shape, px, py): mosaic[px][py] = (tile.id, shape) if len(pstack)==len(tiles): return True result = False if px>0 and not mosaic[px-1][py]: edge_value = shape.W result = dfs_helper(mosaic, graf, px-1, py, pstack, tile, shape.W) elif px<len(mosaic)-1 and not mosaic[px+1][py]: result = dfs_helper(mosaic, graf, px+1, py, pstack, tile, shape.E) elif py<len(mosaic)-1 and not mosaic[px][py+1]: result = dfs_helper(mosaic, graf, px, py+1, pstack, tile, shape.N) # elif py>0 and not mosaic[px][py-1]: # result = dfs_helper(mosaic, graf, px, py-1, pstack, tile, shape.S) if result: return result pstack.pop() mosaic[px][py] = 0 return False data, test_data = load_data() # uncomment the below line to use test data # data = test_data #part one # [id, [lines], [shapes],[matching tiles]] tiles = parse_data(data) edge_len = int(math.sqrt(len(tiles))) graf = defaultdict(dict) for tile in tiles: for shape in tile.shapes: for ival in shape: #todo: iterates over different types. fixit if isinstance(ival,int): graf[ival][tile.id]=tile mosaic = [[0 for _ in range(edge_len)] for _ in range(edge_len)] found = False for i,tile in enumerate(tiles): found = dfs(mosaic, graf, tile,0,0,[]) if found: break if found: result1 = mosaic[0][0][0]\ * mosaic[0][edge_len-1][0]\ * mosaic[edge_len-1][0][0]\ * mosaic[edge_len-1][edge_len-1][0] else: result1 = 'Not found.' print(f'result 1 = {result1}') if result1==13224049461431: print(' -=OK=- ') else: print(' --err-- ') #part two t1 = time.perf_counter() result2 = None #print(f'result 2 = {result2}') t2 = time.perf_counter() #print(f'TIME = {(t2-t1):.2f}')
{ "alphanum_fraction": 0.5621552821, "author": null, "avg_line_length": 27.4069767442, "converted": null, "ext": "py", "file": null, "hexsha": "c63fc3a03d060d3f2f77d8c1204846eb72c77a85", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "82558aeb073efd475943da0040bbd538e7f75eff", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "pascalddg/AoC", "max_forks_repo_path": "AoC2020/aoc20-20-code.py", "max_issues_count": 4, "max_issues_repo_head_hexsha": "82558aeb073efd475943da0040bbd538e7f75eff", "max_issues_repo_issues_event_max_datetime": "2021-01-06T22:02:28.000Z", "max_issues_repo_issues_event_min_datetime": "2021-01-06T10:19:10.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "pascalddg/AoC", "max_issues_repo_path": "AoC2020/aoc20-20-code.py", "max_line_length": 88, "max_stars_count": null, "max_stars_repo_head_hexsha": "82558aeb073efd475943da0040bbd538e7f75eff", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "pascalddg/AoC", "max_stars_repo_path": "AoC2020/aoc20-20-code.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1328, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4714 }
from kivy.config import Config # Config.set('kivy', 'exit_on_escape', '0') # Config.set('graphics', 'resizable', '0') Config.set('graphics', 'width', '640') Config.set('graphics', 'height', '480') import os import cv2 from detection import Face_Detector, Landmark_Detector from faceswap_cam import face_swap from kivy.app import App from kivy.lang import Builder from kivy.clock import Clock from kivy.uix.label import Label from kivy.uix.dropdown import DropDown from kivy.uix.button import Button from kivy.uix.screenmanager import Screen, ScreenManager # from kivy.properties import ObjectProperty import numpy as np class MyScreenManager(ScreenManager): pass class PreScreen(Screen): pass class FrontScreen(Screen): def __init__(self, **kwargs): super(FrontScreen, self).__init__(**kwargs) self.refresh_dt = 0.05 def on_enter(self, *args): # only works for multiple screens? self.face_detector = Face_Detector() self.lmk_detector = Landmark_Detector() self.portraits = os.listdir('./portraits/') # print(self.portraits) ''' dropdown menu ''' self.dropdown = DropDown() for face in self.portraits: btn = Button(text=face, size_hint_y=None, height=32, # color = (0,0,1,1), # background_normal='',background_color=(0.11, 0.17, 0.44, 0.2) ) btn.bind(on_release=lambda btn: self.dropdown.select(btn.text)) self.dropdown.add_widget(btn) self.ids.face_selection.bind(on_release=self.dropdown.open) self.dropdown.bind(on_select=lambda instance, x: setattr(self.ids.face_selection, 'text', x)) def initialize(self, target_face): # self.face_detector = Face_Detector() # self.lmk_detector = Landmark_Detector() try: _source = int(self.ids.cam.text) except Exception as ee: _source = self.ids.cam.text self.cap = cv2.VideoCapture( _source ) self.FaceSwap = face_swap( os.path.join('./portraits', target_face) ) def swap_face(self, *args): ret, frame = self.cap.read() frame = cv2.resize(frame, (480,640)) bboxes, _ = self.face_detector.detect(frame) # get faces if len(bboxes) != 0: bbox = bboxes[0] # get the first bbox = bbox.astype(np.int) lmks, PRY_3d = self.lmk_detector.detect(frame, bbox) # get landmarks lmks = lmks.astype(np.int) frame = self.FaceSwap.run(frame,lmks) cv2.imshow("Face Swap", frame) def update(self,*args): Clock.schedule_interval(self.swap_face, self.refresh_dt) def stop(self): Clock.unschedule(self.swap_face) cv2.destroyWindow('Face Swap') root_widget = Builder.load_string(''' MyScreenManager: PreScreen: FrontScreen: <PreScreen>: Image: source: '' allow_stretch: True keep_ratio: False size: root.size Button: text: 'GO' font_size:40 center: root.center color: 1,0,1,1 background_color: 0,0,0,0 on_release: app.root.current = 'front' <FrontScreen>: name: 'front' Image: source: '' allow_stretch: True keep_ratio: False size: root.size Button: id: face_selection center: root.center text: 'Select a face' size: 0.25*root.width, root.height//13 # on_press: print(root.portraits) Label: text: 'Camera' color: (1, 0.6, 0, 1) font_size: 24 center: 0.2*root.width , 0.65*root.height TextInput: id: cam text: '0' font_size: 12 multiline: False center: 0.52*root.width , 0.625*root.height size: (0.3*root.width, root.height//16) padding: [0.02*root.width,self.height // 2 - (self.line_height//2) * len(self._lines), 0, 0] font_size: dp(18) color:(0.11, 0.17, 0.44, 1.0) Button: id: start text: 'START' center: root.width//2, 0.3*root.height height: root.height//13 on_release: root.initialize(face_selection.text) on_release: root.update() Button: id: reset text: 'RESET' center: 1.5*root.width//2, 0.47*root.height height: root.height//13 on_release: root.stop() ''') class faceApp(App): def build(self): self.title = 'Face Swap' return root_widget faceApp().run()
{ "alphanum_fraction": 0.5767217919, "author": null, "avg_line_length": 29.487654321, "converted": null, "ext": "py", "file": null, "hexsha": "d2de28cea11632f26f235750e1c4ad63996da8ba", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 2, "max_forks_repo_forks_event_max_datetime": "2022-02-21T07:10:45.000Z", "max_forks_repo_forks_event_min_datetime": "2022-01-12T15:22:51.000Z", "max_forks_repo_head_hexsha": "c4680fa24d809539507aa49bcd24e22723bfdf3f", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "dykuang/Real-Time-FaceSwap-with-ONNX", "max_forks_repo_path": "faceswap_kivy.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "c4680fa24d809539507aa49bcd24e22723bfdf3f", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "dykuang/Real-Time-FaceSwap-with-ONNX", "max_issues_repo_path": "faceswap_kivy.py", "max_line_length": 104, "max_stars_count": 2, "max_stars_repo_head_hexsha": "c4680fa24d809539507aa49bcd24e22723bfdf3f", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "dykuang/Real-Time-FaceSwap-with-ONNX", "max_stars_repo_path": "faceswap_kivy.py", "max_stars_repo_stars_event_max_datetime": "2022-02-21T07:10:43.000Z", "max_stars_repo_stars_event_min_datetime": "2022-01-12T15:22:49.000Z", "num_tokens": 1198, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4777 }
#include <boost/gil/extension/dynamic_image/algorithm.hpp>
{ "alphanum_fraction": 0.8305084746, "author": null, "avg_line_length": 29.5, "converted": null, "ext": "hpp", "file": null, "hexsha": "c585151cc6fe9f7cccb1e1e59821a4a3ded82ad2", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 4, "max_forks_repo_forks_event_max_datetime": "2021-07-06T03:06:52.000Z", "max_forks_repo_forks_event_min_datetime": "2019-05-28T21:06:37.000Z", "max_forks_repo_head_hexsha": "919621dcd0c157094bed4df752b583ba6ea6409e", "max_forks_repo_licenses": [ "BSL-1.0" ], "max_forks_repo_name": "miathedev/BoostForArduino", "max_forks_repo_path": "src/boost_gil_extension_dynamic_image_algorithm.hpp", "max_issues_count": 2, "max_issues_repo_head_hexsha": "919621dcd0c157094bed4df752b583ba6ea6409e", "max_issues_repo_issues_event_max_datetime": "2021-05-20T23:55:08.000Z", "max_issues_repo_issues_event_min_datetime": "2021-03-26T15:17:35.000Z", "max_issues_repo_licenses": [ "BSL-1.0" ], "max_issues_repo_name": "miathedev/BoostForArduino", "max_issues_repo_path": "src/boost_gil_extension_dynamic_image_algorithm.hpp", "max_line_length": 58, "max_stars_count": 10, "max_stars_repo_head_hexsha": "919621dcd0c157094bed4df752b583ba6ea6409e", "max_stars_repo_licenses": [ "BSL-1.0" ], "max_stars_repo_name": "miathedev/BoostForArduino", "max_stars_repo_path": "src/boost_gil_extension_dynamic_image_algorithm.hpp", "max_stars_repo_stars_event_max_datetime": "2021-07-06T02:48:49.000Z", "max_stars_repo_stars_event_min_datetime": "2018-03-17T00:58:42.000Z", "num_tokens": 14, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 59 }
#!/usr/bin/env python # Copyright 2014-2020 The PySCF Developers. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # # Author: Qiming Sun <osirpt.sun@gmail.com> # ''' Spin-free lambda equation of UHF-CCSD(T) ''' import numpy from pyscf import lib from pyscf.lib import logger from pyscf.cc import ccsd_lambda from pyscf.cc import uccsd_lambda def kernel(mycc, eris=None, t1=None, t2=None, l1=None, l2=None, max_cycle=50, tol=1e-8, verbose=logger.INFO): return ccsd_lambda.kernel(mycc, eris, t1, t2, l1, l2, max_cycle, tol, verbose, make_intermediates, update_lambda) def make_intermediates(mycc, t1, t2, eris): def p6(t): return (t + t.transpose(1,2,0,4,5,3) + t.transpose(2,0,1,5,3,4) + t.transpose(0,2,1,3,5,4) + t.transpose(2,1,0,5,4,3) + t.transpose(1,0,2,4,3,5)) def r6(w): return (w + w.transpose(2,0,1,3,4,5) + w.transpose(1,2,0,3,4,5) - w.transpose(2,1,0,3,4,5) - w.transpose(0,2,1,3,4,5) - w.transpose(1,0,2,3,4,5)) imds = uccsd_lambda.make_intermediates(mycc, t1, t2, eris) t1a, t1b = t1 t2aa, t2ab, t2bb = t2 nocca, noccb = t2ab.shape[:2] mo_ea, mo_eb = eris.mo_energy eia = mo_ea[:nocca,None] - mo_ea[nocca:] eIA = mo_eb[:noccb,None] - mo_eb[noccb:] fvo = eris.focka[nocca:,:nocca] fVO = eris.fockb[noccb:,:noccb] # aaa d3 = lib.direct_sum('ia+jb+kc->ijkabc', eia, eia, eia) w = numpy.einsum('ijae,kceb->ijkabc', t2aa, numpy.asarray(eris.get_ovvv()).conj()) w-= numpy.einsum('mkbc,iajm->ijkabc', t2aa, numpy.asarray(eris.ovoo.conj())) v = numpy.einsum('jbkc,ia->ijkabc', numpy.asarray(eris.ovov).conj(), t1a) v+= numpy.einsum('jkbc,ai->ijkabc', t2aa, fvo) * .5 rw = r6(p6(w)) / d3 imds.l1a_t = numpy.einsum('ijkabc,jbkc->ia', rw.conj(), numpy.asarray(eris.ovov)) / eia * .25 wvd = r6(p6(w * 2 + v)) / d3 l2_t = numpy.einsum('ijkabc,kceb->ijae', wvd, numpy.asarray(eris.get_ovvv()).conj()) l2_t -= numpy.einsum('ijkabc,iajm->mkbc', wvd, numpy.asarray(eris.ovoo.conj())) l2_t = l2_t + l2_t.transpose(1,0,3,2) l2_t += numpy.einsum('ijkabc,ai->jkbc', rw, fvo) imds.l2aa_t = l2_t.conj() / lib.direct_sum('ia+jb->ijab', eia, eia) * .5 # bbb d3 = lib.direct_sum('ia+jb+kc->ijkabc', eIA, eIA, eIA) w = numpy.einsum('ijae,kceb->ijkabc', t2bb, numpy.asarray(eris.get_OVVV()).conj()) w-= numpy.einsum('imab,kcjm->ijkabc', t2bb, numpy.asarray(eris.OVOO.conj())) v = numpy.einsum('jbkc,ia->ijkabc', numpy.asarray(eris.OVOV).conj(), t1b) v+= numpy.einsum('jkbc,ai->ijkabc', t2bb, fVO) * .5 rw = r6(p6(w)) / d3 imds.l1b_t = numpy.einsum('ijkabc,jbkc->ia', rw.conj(), numpy.asarray(eris.OVOV)) / eIA * .25 wvd = r6(p6(w * 2 + v)) / d3 l2_t = numpy.einsum('ijkabc,kceb->ijae', wvd, numpy.asarray(eris.get_OVVV()).conj()) l2_t -= numpy.einsum('ijkabc,iajm->mkbc', wvd, numpy.asarray(eris.OVOO.conj())) l2_t = l2_t + l2_t.transpose(1,0,3,2) l2_t += numpy.einsum('ijkabc,ai->jkbc', rw, fVO) imds.l2bb_t = l2_t.conj() / lib.direct_sum('ia+jb->ijab', eIA, eIA) * .5 # baa def r4(w): w = w - w.transpose(0,2,1,3,4,5) w = w + w.transpose(0,2,1,3,5,4) return w d3 = lib.direct_sum('ia+jb+kc->ijkabc', eIA, eia, eia) w = numpy.einsum('jIeA,kceb->IjkAbc', t2ab, numpy.asarray(eris.get_ovvv()).conj()) * 2 w += numpy.einsum('jIbE,kcEA->IjkAbc', t2ab, numpy.asarray(eris.get_ovVV()).conj()) * 2 w += numpy.einsum('jkbe,IAec->IjkAbc', t2aa, numpy.asarray(eris.get_OVvv()).conj()) w -= numpy.einsum('mIbA,kcjm->IjkAbc', t2ab, numpy.asarray(eris.ovoo).conj()) * 2 w -= numpy.einsum('jMbA,kcIM->IjkAbc', t2ab, numpy.asarray(eris.ovOO).conj()) * 2 w -= numpy.einsum('jmbc,IAkm->IjkAbc', t2aa, numpy.asarray(eris.OVoo).conj()) v = numpy.einsum('jbkc,IA->IjkAbc', numpy.asarray(eris.ovov).conj(), t1b) v += numpy.einsum('kcIA,jb->IjkAbc', numpy.asarray(eris.ovOV).conj(), t1a) v += numpy.einsum('kcIA,jb->IjkAbc', numpy.asarray(eris.ovOV).conj(), t1a) v += numpy.einsum('jkbc,AI->IjkAbc', t2aa, fVO) * .5 v += numpy.einsum('kIcA,bj->IjkAbc', t2ab, fvo) * 2 rw = r4(w) / d3 imds.l1a_t += numpy.einsum('ijkabc,kcia->jb', rw.conj(), numpy.asarray(eris.ovOV)) / eia * .5 imds.l1b_t += numpy.einsum('ijkabc,jbkc->ia', rw.conj(), numpy.asarray(eris.ovov)) / eIA * .25 wvd = r4(w * 2 + v) / d3 l2_t = numpy.einsum('ijkabc,iaec->jkbe', wvd, numpy.asarray(eris.get_OVvv()).conj()) l2_t -= numpy.einsum('ijkabc,iakm->jmbc', wvd, numpy.asarray(eris.OVoo).conj()) l2_t = l2_t + l2_t.transpose(1,0,3,2) l2_t += numpy.einsum('ijkabc,ai->jkbc', rw, fVO) imds.l2aa_t += l2_t.conj() / lib.direct_sum('ia+jb->ijab', eia, eia) * .5 l2_t = numpy.einsum('ijkabc,kceb->jiea', wvd, numpy.asarray(eris.get_ovvv()).conj()) l2_t += numpy.einsum('ijkabc,kcea->jibe', wvd, numpy.asarray(eris.get_ovVV()).conj()) l2_t -= numpy.einsum('ijkabc,kcjm->miba', wvd, numpy.asarray(eris.ovoo).conj()) l2_t -= numpy.einsum('ijkabc,kcim->jmba', wvd, numpy.asarray(eris.ovOO).conj()) l2_t += numpy.einsum('ijkabc,bj->kica', rw, fvo) imds.l2ab_t = l2_t.conj() / lib.direct_sum('ia+jb->ijab', eia, eIA) * .5 # bba d3 = lib.direct_sum('ia+jb+kc->ijkabc', eia, eIA, eIA) w = numpy.einsum('ijae,kceb->ijkabc', t2ab, numpy.asarray(eris.get_OVVV()).conj()) * 2 w += numpy.einsum('ijeb,kcea->ijkabc', t2ab, numpy.asarray(eris.get_OVvv()).conj()) * 2 w += numpy.einsum('jkbe,iaec->ijkabc', t2bb, numpy.asarray(eris.get_ovVV()).conj()) w -= numpy.einsum('imab,kcjm->ijkabc', t2ab, numpy.asarray(eris.OVOO).conj()) * 2 w -= numpy.einsum('mjab,kcim->ijkabc', t2ab, numpy.asarray(eris.OVoo).conj()) * 2 w -= numpy.einsum('jmbc,iakm->ijkabc', t2bb, numpy.asarray(eris.ovOO).conj()) v = numpy.einsum('jbkc,ia->ijkabc', numpy.asarray(eris.OVOV).conj(), t1a) v += numpy.einsum('iakc,jb->ijkabc', numpy.asarray(eris.ovOV).conj(), t1b) v += numpy.einsum('iakc,jb->ijkabc', numpy.asarray(eris.ovOV).conj(), t1b) v += numpy.einsum('JKBC,ai->iJKaBC', t2bb, fvo) * .5 v += numpy.einsum('iKaC,BJ->iJKaBC', t2ab, fVO) * 2 rw = r4(w) / d3 imds.l1a_t += numpy.einsum('ijkabc,jbkc->ia', rw.conj(), numpy.asarray(eris.OVOV)) / eia * .25 imds.l1b_t += numpy.einsum('ijkabc,iakc->jb', rw.conj(), numpy.asarray(eris.ovOV)) / eIA * .5 wvd = r4(w * 2 + v) / d3 l2_t = numpy.einsum('ijkabc,iaec->jkbe', wvd, numpy.asarray(eris.get_ovVV()).conj()) l2_t -= numpy.einsum('ijkabc,iakm->jmbc', wvd, numpy.asarray(eris.ovOO).conj()) l2_t = l2_t + l2_t.transpose(1,0,3,2) l2_t += numpy.einsum('ijkabc,ai->jkbc', rw, fvo) imds.l2bb_t += l2_t.conj() / lib.direct_sum('ia+jb->ijab', eIA, eIA) * .5 l2_t = numpy.einsum('ijkabc,kceb->ijae', wvd, numpy.asarray(eris.get_OVVV()).conj()) l2_t += numpy.einsum('ijkabc,kcea->ijeb', wvd, numpy.asarray(eris.get_OVvv()).conj()) l2_t -= numpy.einsum('ijkabc,kcjm->imab', wvd, numpy.asarray(eris.OVOO).conj()) l2_t -= numpy.einsum('ijkabc,kcim->mjab', wvd, numpy.asarray(eris.OVoo).conj()) l2_t += numpy.einsum('ijkabc,bj->ikac', rw, fVO) imds.l2ab_t += l2_t.conj() / lib.direct_sum('ia+jb->ijab', eia, eIA) * .5 return imds def update_lambda(mycc, t1, t2, l1, l2, eris=None, imds=None): if eris is None: eris = mycc.ao2mo() if imds is None: imds = make_intermediates(mycc, t1, t2, eris) l1, l2 = uccsd_lambda.update_lambda(mycc, t1, t2, l1, l2, eris, imds) l1a, l1b = l1 l2aa, l2ab, l2bb = l2 l1a += imds.l1a_t l1b += imds.l1b_t l2aa += imds.l2aa_t l2ab += imds.l2ab_t l2bb += imds.l2bb_t return (l1a, l1b), (l2aa, l2ab, l2bb) if __name__ == '__main__': from pyscf import gto from pyscf import scf from pyscf import cc mol = gto.Mole() mol.atom = [ [8 , (0. , 0. , 0.)], [1 , (0. , -0.757 , 0.587)], [1 , (0. , 0.757 , 0.587)]] mol.basis = '631g' mol.spin = 0 mol.build() mf0 = mf = scf.RHF(mol).run(conv_tol=1) mf = scf.addons.convert_to_uhf(mf) mycc = cc.UCCSD(mf) eris = mycc.ao2mo() from pyscf.cc import ccsd_t_lambda_slow as ccsd_t_lambda mycc0 = cc.CCSD(mf0) eris0 = mycc0.ao2mo() mycc0.kernel(eris=eris0) t1 = mycc0.t1 t2 = mycc0.t2 imds = ccsd_t_lambda.make_intermediates(mycc0, t1, t2, eris0) l1, l2 = ccsd_t_lambda.update_lambda(mycc0, t1, t2, t1, t2, eris0, imds) l1ref, l2ref = ccsd_t_lambda.update_lambda(mycc0, t1, t2, l1, l2, eris0, imds) t1 = (t1, t1) t2aa = t2 - t2.transpose(1,0,2,3) t2 = (t2aa, t2, t2aa) l1 = (l1, l1) l2aa = l2 - l2.transpose(1,0,2,3) l2 = (l2aa, l2, l2aa) imds = make_intermediates(mycc, t1, t2, eris) l1, l2 = update_lambda(mycc, t1, t2, l1, l2, eris, imds) print(abs(l2[1]-l2[1].transpose(1,0,2,3)-l2[0]).max()) print(abs(l2[1]-l2[1].transpose(0,1,3,2)-l2[2]).max()) print(abs(l1[0]-l1ref).max()) print(abs(l2[1]-l2ref).max())
{ "alphanum_fraction": 0.6080818414, "author": null, "avg_line_length": 45.4651162791, "converted": null, "ext": "py", "file": null, "hexsha": "86e1293f2c61d1a13edf3dcf1e7a5ed2f5b43296", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 273, "max_forks_repo_forks_event_max_datetime": "2022-03-30T12:25:28.000Z", "max_forks_repo_forks_event_min_datetime": "2018-11-26T10:10:24.000Z", "max_forks_repo_head_hexsha": "0ed03633b699505c7278f1eb501342667d0aa910", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "QuESt-Calculator/pyscf", "max_forks_repo_path": "pyscf/cc/uccsd_t_lambda.py", "max_issues_count": 710, "max_issues_repo_head_hexsha": "0ed03633b699505c7278f1eb501342667d0aa910", "max_issues_repo_issues_event_max_datetime": "2022-03-30T03:53:12.000Z", "max_issues_repo_issues_event_min_datetime": "2018-11-26T22:04:52.000Z", "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "QuESt-Calculator/pyscf", "max_issues_repo_path": "pyscf/cc/uccsd_t_lambda.py", "max_line_length": 91, "max_stars_count": 501, "max_stars_repo_head_hexsha": "0ed03633b699505c7278f1eb501342667d0aa910", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "QuESt-Calculator/pyscf", "max_stars_repo_path": "pyscf/cc/uccsd_t_lambda.py", "max_stars_repo_stars_event_max_datetime": "2022-03-31T11:53:18.000Z", "max_stars_repo_stars_event_min_datetime": "2018-12-06T23:48:17.000Z", "num_tokens": 3910, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 9775 }
import logging from typing import Dict from typing import Sequence from typing import Tuple import yaml import argparse import torch import torch.nn.functional as F from espnet2.tts.abs_tts import AbsTTS from espnet.nets.pytorch_backend.gradtts.diffusion import Diffusion from espnet2.torch_utils.device_funcs import force_gatherable from espnet2.tts.fastspeech2 import FastSpeech2 from espnet.nets.pytorch_backend.nets_utils import make_non_pad_mask class GradFastSpeech2(AbsTTS): def __init__( self, idim, odim, config_file=None, model_file=None, ddim: int = 64, beta_min: float = 0.05, beta_max: float = 20.0, pe_scale: int = 1000, ) -> None: super().__init__() with open(config_file, "r", encoding="utf-8") as f: args = yaml.safe_load(f) args = argparse.Namespace(**args) self.idim = idim self.odim = odim self.padding_idx = 0 self.eos = idim - 1 self.fastspeech2 = FastSpeech2(idim=len(args.token_list), odim=odim, **args.tts_conf) tmp = torch.load(model_file) d = {} for key, value in tmp.items(): if key.startswith("tts"): d[key[4:]] = value self.fastspeech2.load_state_dict(d) for p in self.fastspeech2.parameters(): p.requires_grad = False self.diffusion = Diffusion(ddim, beta_min, beta_max, pe_scale) self.criterion = GradFastSpeech2Loss(odim) def forward( self, text: torch.Tensor, text_lengths: torch.Tensor, speech: torch.Tensor, speech_lengths: torch.Tensor, durations: torch.Tensor, durations_lengths: torch.Tensor, pitch: torch.Tensor, pitch_lengths: torch.Tensor, energy: torch.Tensor, energy_lengths: torch.Tensor, spembs: torch.Tensor = None, ) -> Tuple[torch.Tensor, Dict[str, torch.Tensor], torch.Tensor]: text = text[:, : text_lengths.max()] # for data-parallel speech = speech[:, : speech_lengths.max()] # for data-parallel durations = durations[:, : durations_lengths.max()] # for data-parallel pitch = pitch[:, : pitch_lengths.max()] # for data-parallel energy = energy[:, : energy_lengths.max()] # for data-parallel batch_size = text.size(0) # Add eos at the last of sequence xs = F.pad(text, [0, 1], "constant", self.padding_idx) for i, l in enumerate(text_lengths): xs[i, l] = self.eos ilens = text_lengths + 1 ys, ds, ps, es = speech, durations, pitch, energy olens = speech_lengths before_outs, after_outs, d_outs, p_outs, e_outs = self.fastspeech2._forward( xs, ilens, ys, olens, ds, ps, es, spembs=spembs, is_inference=False ) ys = speech.transpose(1, 2) y_masks = self._source_mask(olens) mu = after_outs.transpose(1, 2) if ys.size(2) % 4 != 0: ys = torch.cat([ys, torch.zeros([batch_size, self.odim, 4 - ys.size(2) % 4], dtype=ys.dtype, device=ys.device)], dim=2) mu = torch.cat([mu, torch.zeros([mu.size(0), self.odim, 4 - mu.size(2) % 4], dtype=mu.dtype, device=mu.device)], dim=2) y_masks = torch.cat([y_masks, torch.zeros([y_masks.size(0), 1, 4 - y_masks.size(2) % 4], dtype=y_masks.dtype, device=y_masks.device)], dim=2) noise_estimation, z = self.diffusion(ys, y_masks, mu) diff_loss = self.criterion(noise_estimation, z, y_masks) loss = diff_loss stats = dict( diff_loss=diff_loss.item(), loss=loss.item(), ) loss, stats, weight = force_gatherable((loss, stats, batch_size), loss.device) return loss, stats, weight def inference( self, text: torch.Tensor, timesteps: int = 50, spembs: torch.Tensor = None, temperature: float = 1.0, alpha: float = 1.03, use_teacher_forcing: bool = False, ): mu, _, _ = self.fastspeech2.inference(text, alpha=alpha, use_teacher_forcing=use_teacher_forcing) # import numpy as np # np.save("/nolan/inference/gradtts_pre.mel.npy", mu.data.cpu().numpy()) length = mu.shape[0] olens = torch.tensor([length], dtype=torch.long, device=mu.device) y_masks = self._source_mask(olens) mu = mu.unsqueeze(0).transpose(1, 2) if mu.size(2) % 4 != 0: mu = torch.cat([mu, torch.zeros([1, self.odim, 4 - mu.size(2) % 4], dtype=mu.dtype, device=mu.device)], dim=2) y_masks = torch.cat([y_masks, torch.zeros([1, 1, 4 - y_masks.size(2) % 4], dtype=y_masks.dtype, device=y_masks.device)], dim=2) z = mu + torch.randn_like(mu, device=mu.device) / temperature out = self.diffusion.inference(z, y_masks, mu, timesteps).transpose(1, 2) return out[0, :length, :], None, None def _source_mask(self, ilens: torch.Tensor) -> torch.Tensor: x_masks = make_non_pad_mask(ilens).to(next(self.parameters()).device) return x_masks.unsqueeze(-2) class GradFastSpeech2Loss(torch.nn.Module): def __init__(self, odim): super().__init__() self.odim = odim def forward( self, noise_estimation: torch.Tensor, z: torch.Tensor, y_masks: torch.Tensor, ): diff_loss = torch.sum((noise_estimation + z) ** 2) / (torch.sum(y_masks) * self.odim) return diff_loss
{ "alphanum_fraction": 0.6089778259, "author": null, "avg_line_length": 37.2281879195, "converted": null, "ext": "py", "file": null, "hexsha": "213686a082af73aa65210e99ccede9eeda2ab854", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-10-18T08:15:31.000Z", "max_forks_repo_forks_event_min_datetime": "2021-10-18T08:15:31.000Z", "max_forks_repo_head_hexsha": "73c43ac034d7a44a3bf1a10dc4079884df57de43", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "NolanYuu/espnet", "max_forks_repo_path": "espnet2/tts/gradfastspeech2.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "73c43ac034d7a44a3bf1a10dc4079884df57de43", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "NolanYuu/espnet", "max_issues_repo_path": "espnet2/tts/gradfastspeech2.py", "max_line_length": 153, "max_stars_count": null, "max_stars_repo_head_hexsha": "73c43ac034d7a44a3bf1a10dc4079884df57de43", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "NolanYuu/espnet", "max_stars_repo_path": "espnet2/tts/gradfastspeech2.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1495, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 5547 }
# quantum solver import numpy as np from scipy.linalg import eigh def solver(nk, nbasis, mu, hamiton_mat): kpoints = np.linspace(0, np.pi, nk) # build and solve eigenvalue problem T = 0 # En = np.zeros((nk, nbasis)) density = np.zeros(nbasis, dtype=np.complex64) for ki, k in enumerate(kpoints): kinetic_term = np.array([0.5*(k+(i-nbasis//2)*2*np.pi)**2 for i in range(nbasis)]) np.fill_diagonal(hamiton_mat, kinetic_term) En_k, Uq_k = eigh(hamiton_mat, overwrite_a=True, overwrite_b=True) # En[ki] = En_k # compute mu if k == 0: b = En_k[0] # set the minimum of band energy to 0 ! # compute electron density # compute kinetic energy num_mat_eigspace = np.zeros((nbasis, nbasis)) for i in range(nbasis): if En_k[i] <= mu + b: num_mat_eigspace[i, i] = 1 else: break density_mat_kspace = Uq_k @ (num_mat_eigspace @ (Uq_k.T).conj()) density_k = np.zeros(nbasis, dtype=np.complex64) T_k = 0 for i in range(nbasis): density_k[i] = np.trace(density_mat_kspace, offset=i) T_k += 0.5*((k+(i-nbasis//2)*2*np.pi)**2)*(density_mat_kspace[i, i]).real T += T_k density += density_k return T/nk, density/nk
{ "alphanum_fraction": 0.5734317343, "author": null, "avg_line_length": 33.0487804878, "converted": null, "ext": "py", "file": null, "hexsha": "6fc2f97757e6319b9f7efd431ff80db0d256bc74", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "a8ebd46ec0442a1fcbfa7e0571f43550e1faaa70", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "HamletWantToCode/machine_learning_kinetic_energy", "max_forks_repo_path": "main/solver.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "a8ebd46ec0442a1fcbfa7e0571f43550e1faaa70", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "HamletWantToCode/machine_learning_kinetic_energy", "max_issues_repo_path": "main/solver.py", "max_line_length": 90, "max_stars_count": 2, "max_stars_repo_head_hexsha": "a8ebd46ec0442a1fcbfa7e0571f43550e1faaa70", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "HamletWantToCode/machine_learning_kinetic_energy", "max_stars_repo_path": "main/solver.py", "max_stars_repo_stars_event_max_datetime": "2019-02-24T09:22:37.000Z", "max_stars_repo_stars_event_min_datetime": "2018-12-17T01:11:21.000Z", "num_tokens": 401, "path": null, "reason": "import numpy,from scipy", "repo": null, "save_path": null, "sha": null, "size": 1355 }
import oemof.solph as solph from .component_electrolyzer import Electrolyzer import pyomo.environ as po class ElectrolyzerWasteHeat(Electrolyzer): """ Electrolyzer agents with waste heat model are created through this subclass of the Electrolyzer class """ def __init__(self, params): # Split the params dict param_bus_th = {"bus_th": params.pop("bus_th")} # Call the init function of the mother class. Electrolyzer.__init__(self, params) """ PARAMETERS """ # Define the additional thermal bus self.bus_th = None """ UPDATE PARAMETER DEFAULT VALUES """ self.set_parameters(param_bus_th) # Interval time [min]. self.interval_time = self.sim_params.interval_time # Calculate the max. energy the electrolyzer can use in one time step [Wh]. self.energy_max = self.power_max * self.interval_time / 60 """ CONSTANT PARAMETERS (PHYSICS) """ # constant parameters for calculating sensible heat: # specific heat at constant pressure [J/(kg*K)] self.c_p_H2 = 14304 self.c_p_O2 = 920 self.c_p_H2O = 4183 """ ELECTROLYZER GEOMETRY PARAMETERS """ self.diameter_cell = (4 * self.area_cell / 3.14) ** 0.5 / 100 # m # The height of the end of the stack which is not part of the cells is assumed to have a # dependence on the diameter of the cell. The ratio is taken as 7 : 120 # (stack_end_height : diameter_cell), which is based on De Silva, Y.S.K. (2017). Design # of an Alkaline Electrolysis Stack, University of Agder. self.stack_end_height = 0.058 * self.diameter_cell # The height of an individual cell in relation to cell diameter is calculated using example # data from Vogt, U.F. et al. (2014). Novel Developments in Alkaline Water Electrolysis, # Empa Laboratory of Hydrogen and Energy. The individual cell height is estimated and # compared with the given cell radius, and a ratio of 1 : 75.5 is obtained. self.height_cell = self.diameter_cell / 75.5 # The total stack height is calculated by taking the cell stack and the two ends of the # stack into consideration self.height_stack = (self.height_cell * self.z_cell) + (2 * self.stack_end_height) # The external surface area of the electrolysis stack is calculated assuming that it is # cylindrical self.area_stack = ( 2 * self.area_cell / 10000 + 3.14 * self.diameter_cell * self.height_stack ) # [m^2] # The overall surface area exposed by the gas separators and the pipe communicating # them is assumed to be in a ratio of 1 : 0.42 with the area of the stack (taken from # Dieguez et al) self.area_separator = 2.38 * self.area_stack # Save the two models to set constraints later. self.model_h2 = None self.model_th = None def conversion_fun_ely(self, ely_energy): # Create a function that will give out the mass values for the electric energy # values at the breakpoints. # Check the index of this ely_energy entry. this_index = self.supporting_points["energy_halved"].index(ely_energy) # Return the according hydrogen production value [kg]. return self.supporting_points["h2_produced"][this_index] def conversion_fun_thermal(self, ely_energy): # Create a function that will give out the thermal energy values for the electric # energy values at the breakpoints. # Check the index of this ely_energy entry. this_index = self.supporting_points["energy_halved"].index(ely_energy) # Return the according hydrogen production value [kg]. return self.supporting_points["thermal_energy"][this_index] def create_oemof_model(self, busses, model): # Get the non-linear behaviour. self.update_nonlinear_behaviour() # First create the hydrogen producing oemof component electrolyzer = solph.custom.PiecewiseLinearTransformer( label=self.name, inputs={ busses[self.bus_el]: solph.Flow( nominal_value=self.energy_max / 2, variable_costs=0 ) }, outputs={busses[self.bus_h2]: solph.Flow()}, in_breakpoints=self.supporting_points["energy_halved"], conversion_function=self.conversion_fun_ely, pw_repn="CC", ) # Then create the thermal oemof component. electrolyzer_thermal = solph.custom.PiecewiseLinearTransformer( label=self.name + "_thermal", inputs={ busses[self.bus_el]: solph.Flow( nominal_value=self.energy_max / 2, variable_costs=0 ) }, outputs={busses[self.bus_th]: solph.Flow()}, in_breakpoints=self.supporting_points["energy_halved"], conversion_function=self.conversion_fun_thermal, pw_repn="CC", ) # Add the two components to the model. model.add(electrolyzer, electrolyzer_thermal) self.model_h2 = electrolyzer self.model_th = electrolyzer_thermal return None def update_nonlinear_behaviour(self): # Set up the breakpoints for the electrolyzer conversion of electricity to hydrogen. n_supporting_point = 10 # Get the breakpoint values for electric energy [Wh] and produced hydrogen [kg]. bp_ely_energy = [] bp_ely_h2 = [] bp_ely_temp = [] bp_ely_thermal = [] for i_supporting_point in range(n_supporting_point + 1): # Calculate the energy for this breakpoint [Wh]. this_energy = ( i_supporting_point / n_supporting_point * self.energy_max ) bp_ely_energy.append(this_energy) # Calculate the hydrogen produced [kg] and resulting temperature [K] with the # energy of this breakpoint and at the current temperature. [this_mass, this_temp] = self.get_mass_and_temp(this_energy / 1000) bp_ely_h2.append(this_mass) bp_ely_temp.append(this_temp) # Calculate the waste heat [Wh] with the energy, hydrogen produced and resulting # temperature of this breakpoint at the current temperature. this_waste_heat = ( self.get_waste_heat(this_energy / 1000, this_mass, this_temp) * 1000 ) # [Wh] bp_ely_thermal.append(this_waste_heat) self.supporting_points["temperature"] = bp_ely_temp self.supporting_points["h2_produced"] = bp_ely_h2 self.supporting_points["energy"] = bp_ely_energy self.supporting_points["thermal_energy"] = bp_ely_thermal self.supporting_points["energy_halved"] = [ this_bp / 2 for this_bp in bp_ely_energy ] def get_waste_heat(self, energy_used, h2_produced, new_ely_temp): # source: Dieguez et al., 'Thermal Performance of a commercial alkaline # water electrolyzer: Experimental study and mathematical modeling', # Int. J. Hydrogen Energy, 2008 energy_used [kWh] --> internal_heat_generation [kWh] internal_heat_generation = ( energy_used - h2_produced * self.upp_heat_val * 1e6 / 3600 / 1000 ) # [kWh] # heat losses: dT = new_ely_temp - self.temp_min # [K] # equation from Dieguez et al: heat_transfer_coefficient = ( 1.32 * (dT / self.diameter_cell) ** 0.25 ) # [W/(m^2*K)] heat_losses = ( heat_transfer_coefficient * (self.area_stack + self.area_separator) * dT * self.interval_time / 60 / 1000) # [kWh] [sensible_heat, latent_heat] = self.sensible_and_latent_heats( h2_produced, new_ely_temp ) # [kWh] if new_ely_temp >= (0.999 * self.temp_max): waste_heat = internal_heat_generation - heat_losses + sensible_heat else: waste_heat = 0 return waste_heat def sensible_and_latent_heats(self, mass_H2, new_ely_temp): # mass of H2, O2 and H2O is related by the water decomposition stoichiometry # and the mass balance mass_O2 = mass_H2 * 0.5 * self.molar_mass_O2 / self.molarity # as a first approximation mass_H2O_vapor is neglected in the mass balance, # since condensers temperature and pressure are not known mass_H2O = mass_H2 + mass_O2 # sensible heat removed from the system with the H2 and O2 streams, as well # as the sensible heat required to warm the deionized water from room temperature # to the stack operating temperature sensible_heat = ( mass_H2O * self.c_p_H2O * (self.temp_min - new_ely_temp) - mass_H2 * self.c_p_H2 * (new_ely_temp - self.temp_min) - mass_O2 * self.c_p_O2 * (new_ely_temp - self.temp_min) ) / 3.6e6 # [kWh], 1J = 1/3.6e6 kWh # latent heat is neglected since mass_H2O_vapor is neglected latent_heat = 0 return [sensible_heat, latent_heat] def update_constraints(self, busses, model_to_solve): # Set a constraint so that the electric inflow of the hydrogen producing and the # thermal part are always the same (which is necessary while the piecewise linear # transformer cannot have two outputs yet and therefore the two parts need to be # separate components). def electrolyzer_ratio_rule(model, t): # Inverter flow expr = 0 expr += model.flow[busses[self.bus_el], self.model_th, t] # force discharge to zero when grid available expr += -model.flow[busses[self.bus_el], self.model_h2, t] return expr == 0 model_to_solve.electrolyzer_flow_ratio_fix = po.Constraint( model_to_solve.TIMESTEPS, rule=electrolyzer_ratio_rule ) def update_flows(self, results, sim_params): # Check if the component has an attribute 'flows', if not, create it as an empty dict. Electrolyzer.update_flows(self, results, sim_params, self.name) Electrolyzer.update_flows( self, results, sim_params, self.name + "_thermal" )
{ "alphanum_fraction": 0.6440026889, "author": null, "avg_line_length": 46.0752212389, "converted": null, "ext": "py", "file": null, "hexsha": "1f1ae7c604af497219008ebfbf892eec12275e4a", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "06f0c9a6af51be73ef1d5eaaeed9946be672e51e", "max_forks_repo_licenses": [ "ECL-2.0", "Apache-2.0", "MIT-0", "MIT" ], "max_forks_repo_name": "morrme/smooth", "max_forks_repo_path": "smooth/components/component_electrolyzer_waste_heat.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "06f0c9a6af51be73ef1d5eaaeed9946be672e51e", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "ECL-2.0", "Apache-2.0", "MIT-0", "MIT" ], "max_issues_repo_name": "morrme/smooth", "max_issues_repo_path": "smooth/components/component_electrolyzer_waste_heat.py", "max_line_length": 99, "max_stars_count": null, "max_stars_repo_head_hexsha": "06f0c9a6af51be73ef1d5eaaeed9946be672e51e", "max_stars_repo_licenses": [ "ECL-2.0", "Apache-2.0", "MIT-0", "MIT" ], "max_stars_repo_name": "morrme/smooth", "max_stars_repo_path": "smooth/components/component_electrolyzer_waste_heat.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2466, "path": null, "reason": "import pyomo", "repo": null, "save_path": null, "sha": null, "size": 10413 }
from efficientnet_pytorch import EfficientNet import torch from torch import nn from torch import optim from torch.utils.data import TensorDataset, DataLoader, Dataset import tokenizers from torchvision import transforms from PIL import Image from tqdm import tqdm import numpy as np import pandas as pd model = EfficientNet.from_pretrained('efficientnet-b7', advprop=False) device = torch.device("cuda" if torch.cuda.is_available() else "cpu") cpu = torch.device('cpu') print(f"使用デバイス: {device}") model.to(device) class ThumbnailDataset(Dataset): def __init__(self, df): self.df = df self.tfms = transforms.Compose( [ transforms.Resize((90, 120)), transforms.ToTensor(), transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]), ] ) def __getitem__(self, index): data = {} row = self.df.iloc[index] data['images'] = self.get_image_data(row) return data def __len__(self): return len(self.df) def get_image_data(self, row): image_tensor = self.tfms(Image.open(f"./data/thumbnail/{row.video_id}.jpg").convert('RGB')) return image_tensor def get_features(df): ds = ThumbnailDataset(df) dl = DataLoader(ds, batch_size=64, shuffle=False) preds = [] for batch in tqdm(dl): input_images = batch["images"] bs = input_images.size(0) output = model.extract_features(input_images) output = nn.AdaptiveAvgPool2d(1)(output) output = output.view(bs, -1) output = output.to(cpu) preds.append(output.detach().clone().numpy()) return np.concatenate(preds, axis=0) def main(): train = pd.read_csv("./data/input/train_data.csv") test = pd.read_csv("./data/input/test_data.csv") train_image_features = get_features(train) test_image_features = get_features(test) num = train_image_features.shape[1] cols = [f"image_{i}" for i in range(num)] train_image_df = pd.DataFrame(train_image_features, columns=cols) test_image_df = pd.DataFrame(test_image_features, columns=cols) train_image_df.to_csv("./data/input/train_image_features.csv", index=False) test_image_df.to_csv("./data/input/test_image_features.csv", index=False)
{ "alphanum_fraction": 0.6686877436, "author": null, "avg_line_length": 31.2027027027, "converted": null, "ext": "py", "file": null, "hexsha": "b8c2869fd005364d7008bc2f6c4240bdf1c8e376", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "04740862fb28fb9a38131554369d6c54eb560fc5", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "Lain-progressivehouse/probspace-youtube", "max_forks_repo_path": "src/efficient_net_features.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "04740862fb28fb9a38131554369d6c54eb560fc5", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "Lain-progressivehouse/probspace-youtube", "max_issues_repo_path": "src/efficient_net_features.py", "max_line_length": 99, "max_stars_count": 5, "max_stars_repo_head_hexsha": "04740862fb28fb9a38131554369d6c54eb560fc5", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "Lain-progressivehouse/probspace-youtube", "max_stars_repo_path": "src/efficient_net_features.py", "max_stars_repo_stars_event_max_datetime": "2021-02-08T03:54:29.000Z", "max_stars_repo_stars_event_min_datetime": "2020-06-29T04:32:07.000Z", "num_tokens": 544, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2309 }
[STATEMENT] lemma sH_seq: "\<^bold>{P\<^bold>} X;Y \<^bold>{Q\<^bold>} = \<^bold>{P\<^bold>} X \<^bold>{\<lambda>s. \<forall>s'. (s, s') \<in> Y \<longrightarrow> Q s'\<^bold>}" [PROOF STATE] proof (prove) goal (1 subgoal): 1. \<^bold>{P\<^bold>}X ; Y\<^bold>{Q\<^bold>} = \<^bold>{P\<^bold>}X\<^bold>{\<lambda>s. \<forall>s'. (s, s') \<in> Y \<longrightarrow> Q s'\<^bold>} [PROOF STEP] unfolding rel_kat_H [PROOF STATE] proof (prove) goal (1 subgoal): 1. (\<forall>s s'. P s \<longrightarrow> (s, s') \<in> X ; Y \<longrightarrow> Q s') = (\<forall>s s'. P s \<longrightarrow> (s, s') \<in> X \<longrightarrow> (\<forall>s'a. (s', s'a) \<in> Y \<longrightarrow> Q s'a)) [PROOF STEP] by auto
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": "Hybrid_Systems_VCs_KleeneAlgebraTests_HS_VC_KAT_rel", "hexsha": null, "include": null, "lang": null, "length": 2, "llama_tokens": 296, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
import numpy as np import pandas as pd import matplotlib.pyplot as plt import re def getting_and_knowing_your_data(chipo): #Step 4. See the first 10 entries print(chipo.head(10)) #Step 5. What is the number of observations in the dataset? print(len(chipo.index)) #Step 6. What is the number of columns in the dataset? print(len(chipo.columns)) #Step 7. Print the name of all the columns. chipoFeatureList= [] [chipoFeatureList.append(feature) for feature in chipo.columns] for feature in chipoFeatureList: print(feature) #Step 8. How is the dataset indexed? print(chipo.info) #Step 9. Which was the most ordered item? #Step 10. How many items were ordered? itemByOrderQuantity = chipo['item_name'].value_counts() print("{} with {} orders".format(itemByOrderQuantity.index[0], itemByOrderQuantity[0])) #Step 11. What was the most ordered item in the choice_description column? choiceDescriptionByQuantity = chipo['choice_description'].value_counts() print("{} with {} orders".format(choiceDescriptionByQuantity.index[0], choiceDescriptionByQuantity[0])) #Step 12. How many items were ordered in total? print(chipo['quantity'].sum()) #Step 13. Turn the item price into a float ( hint: you can create a #function and use the apply method to do this ) cleanItemPrice= [] for i in range(0,len(chipo)): newCell = chipo['item_price'][i] newCell = re.sub('\$', '', newCell) cleanItemPrice.append(float(newCell)) chipo['item_price'].update(cleanItemPrice) chipo['item_price'] = chipo['item_price'].astype(float) print(chipo['item_price'].dtypes) #Step 14. How much was the revenue for the period in the dataset? print(chipo['item_price'].sum(), '$') #Step 15. How many orders were made in the period? print(chipo['order_id'].max()) #Step 16. What is the average amount per order? chipoByOrder = pd.DataFrame(chipo, columns=['order_id', 'item_price']) chipoByOrderMean = chipo.groupby(by='order_id').mean() print(chipoByOrderMean['item_price'].sum()/len(chipoByOrderMean.index)) #Step 17. How many different items are sold? print(len(itemByOrderQuantity.unique())) return chipo def filtering_and_sorting(chipo) #Step 4. How many products cost more than $10.00? #Step 5. What is the price of each item? counterProducts = pd.DataFrame(chipo, columns=['item_name', 'item_price']) counterProductsByType = counterProducts.groupby(by='item_name').mean() totalBelowTenUSD = 0 itemNamePrice = [] for currentPrice in counterProductsByType['item_price']: if currentPrice>10: totalBelowTenUSD += 1 itemNamePrice.append(currentPrice) print('Total items under $10:', totalBelowTenUSD) print('Prices list:', itemNamePrice) #Step 6. Print a data frame with only two columns item_name and item_price print(counterProducts) #Step 7. Sort by the name of the item counterProductsSortedByName = counterProducts.sort_values(by='item_name') print(counterProductsSortedByName) #Step 8. What was the quantity of the most expensive item ordered? print(counterProducts.sort_values(by='item_price').max()) #Step 9. How many times were a Veggie Salad Bowl ordered? counterProductsSortedByName = pd.DataFrame(chipo, columns=['item_name']) counterProductsSortedByName = counterProductsSortedByName.assign(counterValue ='1') counterProductsSortedByName = counterProductsSortedByName.groupby(by='item_name').count() print(counterProductsSortedByName.loc['Veggie Salad Bowl']['counterValue']) #Step 10. How many times people ordered more than one Canned Soda? itemByQuantity = pd.DataFrame(chipo, columns=['item_name', 'quantity']) quantitiesCounter = 0 for i in range(0,len(itemByQuantity)): if itemByQuantity['item_name'][i] == 'Canned Soda' and itemByQuantity['quantity'][i] == 2: quantitiesCounter+=1 print(quantitiesCounter) def grouping(users): #Step 4. Discover what is the mean age per occupation usersMeanAgeByOccupation = pd.DataFrame(users, columns=['age','occupation']) usersMeanAgeByOccupation = usersMeanAgeByOccupation.groupby(by='occupation').mean() print(usersMeanAgeByOccupation) #Step 5. Discover the Male ratio per occupation and sort it from the most to the least usersRatioGenderByOccupation = pd.DataFrame(users, columns=['gender','occupation']) usersRatioGenderByOccupationMale = usersRatioGenderByOccupation.loc[ usersRatioGenderByOccupation['gender'] == 'M'] usersRatioGenderByOccupationFemale = usersRatioGenderByOccupation.loc[ usersRatioGenderByOccupation['gender'] == 'F'] usersRatioGenderByOccupationMale = usersRatioGenderByOccupationMale.groupby( by='occupation').count() usersRatioGenderByOccupationMale = usersRatioGenderByOccupationMale.rename( columns={'gender':'SumMale'}) usersRatioGenderByOccupationFemale = usersRatioGenderByOccupationFemale.groupby( by='occupation').count() usersRatioGenderByOccupationFemale = usersRatioGenderByOccupationFemale.rename( columns={'gender':'SumFemale'}) ratioList = pd.DataFrame(columns=['occupation','ratio']) for i in range(0,len(usersRatioGenderByOccupationMale)): if usersRatioGenderByOccupationMale.index[i] in usersRatioGenderByOccupationFemale.index: currentOccupation = usersRatioGenderByOccupationMale.index[i] ratioValue = usersRatioGenderByOccupationMale['SumMale'][i]/usersRatioGenderByOccupationFemale[ 'SumFemale'][usersRatioGenderByOccupationMale.index[i]] ratioList.loc[len(ratioList.index)] = [currentOccupation,ratioValue] else: currentOccupation = usersRatioGenderByOccupationMale.index[i] ratioValue = 100 ratioList.loc[len(ratioList.index)] = [currentOccupation,ratioValue] for i in range(0,len(usersRatioGenderByOccupationFemale)): if usersRatioGenderByOccupationFemale.index[i] not in usersRatioGenderByOccupationFemale.index: currentOccupation = usersRatioGenderByOccupationFemale.index[i] ratioValue = 0 ratioList.loc[len(ratioList.index)] = [currentOccupation,ratioValue] ratioList = ratioList.sort_values(by = 'ratio', ascending = False) print(ratioList) #Step 6. For each occupation, calculate the minimum and maximum ages usersOccupationAgeMinMax = pd.DataFrame(users, columns=['age','occupation']) occupationListMaxAge = usersOccupationAgeMinMax.groupby(by='occupation').max() occupationListMinAge = usersOccupationAgeMinMax.groupby(by='occupation').min() #Step 7. For each combination of occupation and gender, calculate the mean age usersOccupationGenderMeanAge = pd.DataFrame(users, columns=['occupation', 'gender' ,'age']) usersOccupationGenderMeanAge = usersOccupationGenderMeanAge.groupby(by=['occupation','gender']).mean() #Step 8. For each occupation present the percentage of women and men usersRatioGenderByOccupationSum = usersRatioGenderByOccupationMale[ 'SumMale'] + usersRatioGenderByOccupationFemale['SumFemale'] usersRatioGenderByOccupationSum['doctor']=usersRatioGenderByOccupationMale.loc['doctor'] usersRatioGenderByOccupationRatioMale = (usersRatioGenderByOccupationMale[ 'SumMale'] / usersRatioGenderByOccupationSum) * 100 usersRatioGenderByOccupationRatioFeale = 100 - usersRatioGenderByOccupationRatioMale def merge(data1, data2, data3): #Step 4. Join the two dataframes along rows and assign all_data all_data = pd.concat([data1, data2], ignore_index=True, axis = 0) #Step 5. Join the two dataframes along columns and assign to all_data_col all_data_col = pd.concat([data1, data2], ignore_index=True, axis = 1) #Step 6. Print data3 print(data3) #Step 7. Merge all_data and data3 along the subject_id value all_data_and_data3 = all_data.merge(data3, on = 'subject_id', how = 'right') #Step 8. Merge only the data that has the same 'subject_id' on both data1 and data2 all_data_same = data1.merge(data2, on = 'subject_id', how = 'inner') #Step 9. Merge all values in data1 and data2, with matching records #from both sides where available. all_data_same2 = data1.merge(data2, on = 'subject_id', how = 'outer') def iris_function(iris): #Step 4. Create columns for the dataset iris.columns = ['SepalLength (cm)', 'SepalWidth (cm)', 'PetalLength (cm)', 'PetalWidth (cm)', 'class'] #Step 5. Is there any missing value in the dataframe? irisFeatureList= [] [irisFeatureList.append(feature) for feature in iris.columns] for feature in irisFeatureList: print("{} had {} % missing values".format(feature,np.round(iris[feature].isnull().sum()/len(iris)*100,2))) print ('\n') #Step 6. Let’s set the values of the rows 10 to 29 of the column 'petal_length' to NaN for i in range(10,29): iris.loc[i,'PetalLength (cm)'] = np.NaN #Step 7. Good, now lets substitute the NaN values to 1.0 iris['PetalLength (cm)'] = iris['PetalLength (cm)'].replace(np.nan, 1.0) #Step 8. Now let's delete the column class iris = iris.drop(columns='class') #Step 9. Set the first 3 rows as NaN iris.iloc[0:3,:] = np.nan #Step 10. Delete the rows that have NaN iris = iris.dropna() #Step 11. Reset the index so it begins with 0 again iris = iris.reset_index() return iris def main(): chipo = pd.read_csv('https://raw.githubusercontent.com/justmarkham/DAT8/master/data/chipotle.tsv', sep = '\t') chipo = getting_and_knowing_your_data(chipo) filtering_and_sorting(chipo) users = pd.read_csv('https://raw.githubusercontent.com/justmarkham/DAT8/master/data/u.user', sep = '|') users = grouping(users) data1 = pd.DataFrame({'subject_id': ['1', '2', '3', '4', '5'], 'first_name': ['Alex', 'Amy', 'Allen', 'Alice', 'Ayoung'], 'last_name': ['Anderson', 'Ackerman', 'Ali', 'Aoni', 'Atiches']}) data2 = pd.DataFrame({'subject_id': ['4', '5', '6', '7', '8'], 'first_name': ['Billy', 'Brian', 'Bran', 'Bryce', 'Betty'], 'last_name': ['Bonder', 'Black', 'Balwner', 'Brice', 'Btisan']}) data3 = pd.DataFrame({'subject_id': ['1', '2', '3', '4', '5', '7', '8', '9', '10', '11'], 'test_id': [51, 15, 15, 61, 16, 14, 15, 1, 61, 16]}) merge(data1, data2, data3) iris = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data') iris = iris_function(iris) if __name__ == "__main__": main()
{ "alphanum_fraction": 0.6405439864, "author": null, "avg_line_length": 41.2807017544, "converted": null, "ext": "py", "file": null, "hexsha": "9754575b4fe635780cb9a18b48948eb64925e834", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "71233cf3764b528c39438d5d45b433f094456717", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "ofir-frd/Machine-Learning-Bootcamp", "max_forks_repo_path": "pandas/PandasExerciseBasic.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "71233cf3764b528c39438d5d45b433f094456717", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "ofir-frd/Machine-Learning-Bootcamp", "max_issues_repo_path": "pandas/PandasExerciseBasic.py", "max_line_length": 115, "max_stars_count": null, "max_stars_repo_head_hexsha": "71233cf3764b528c39438d5d45b433f094456717", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "ofir-frd/Machine-Learning-Bootcamp", "max_stars_repo_path": "pandas/PandasExerciseBasic.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2865, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 11765 }
# Copyright (c) Facebook, Inc. and its affiliates. import unittest import torch import random import os import numpy as np from pythia.models.cnn_lstm import CNNLSTM from pythia.common.registry import registry from pythia.common.sample import Sample, SampleList from pythia.utils.configuration import ConfigNode, Configuration from pythia.utils.general import get_pythia_root class TestModelCNNLSTM(unittest.TestCase): def setUp(self): torch.manual_seed(1234) registry.register("clevr_text_vocab_size", 80) registry.register("clevr_num_final_outputs", 32) config_path = os.path.join( get_pythia_root(), "..", "configs", "vqa", "clevr", "cnn_lstm.yml" ) config_path = os.path.abspath(config_path) configuration = Configuration(config_path) configuration.config["datasets"] = "clevr" configuration.freeze() self.config = configuration.config registry.register("config", self.config) def test_forward(self): model_config = self.config.model_attributes.cnn_lstm cnn_lstm = CNNLSTM(model_config) cnn_lstm.build() cnn_lstm.init_losses_and_metrics() self.assertTrue(isinstance(cnn_lstm, torch.nn.Module)) test_sample = Sample() test_sample.text = torch.randint(1, 79, (10,), dtype=torch.long) test_sample.image = torch.randn(3, 320, 480) test_sample.targets = torch.randn(32) test_sample_list = SampleList([test_sample]) test_sample_list.dataset_type = "train" test_sample_list.dataset_name = "clevr" output = cnn_lstm(test_sample_list) scores = output["scores"] loss = output["losses"]["train/logit_bce"] accuracy = output["metrics"]["train/accuracy"] np.testing.assert_almost_equal(loss.item(), 23.4751, decimal=4) np.testing.assert_almost_equal(accuracy.item(), 0) self.assertEqual(scores.size(), torch.Size((1, 32))) expected_scores = [ 2.2298e-02, -2.4975e-01, -1.1960e-01, -5.0868e-01, -9.3013e-02, 1.3202e-02, -1.7536e-01, -3.1180e-01, 1.5369e-01, 1.4900e-01, 1.9006e-01, -1.9457e-01, 1.4924e-02, -1.1032e-01, 1.3777e-01, -3.6255e-01, -2.9327e-01, 5.6247e-04, -4.8732e-01, 4.0949e-01, -1.1069e-01, 2.9696e-01, 4.1903e-02, 6.7062e-02, 7.0094e-01, -1.9898e-01, -2.9502e-03, -3.9040e-01, 1.2218e-01, 3.7895e-02, 2.4472e-02, 1.7213e-01 ] np.testing.assert_almost_equal(scores[0].tolist(), expected_scores, decimal=5)
{ "alphanum_fraction": 0.6508058327, "author": null, "avg_line_length": 38.3235294118, "converted": null, "ext": "py", "file": null, "hexsha": "126d74ed63a92230fa452d77527a1e8fba0cc3b9", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 11, "max_forks_repo_forks_event_max_datetime": "2021-06-24T05:39:36.000Z", "max_forks_repo_forks_event_min_datetime": "2020-03-07T08:10:15.000Z", "max_forks_repo_head_hexsha": "2ad4c4cfe121e951db12efa3fe5b6f1cc0188da7", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "vatsalg29/btp", "max_forks_repo_path": "tests/models/test_cnn_lstm.py", "max_issues_count": 1, "max_issues_repo_head_hexsha": "2ad4c4cfe121e951db12efa3fe5b6f1cc0188da7", "max_issues_repo_issues_event_max_datetime": "2020-10-05T10:11:24.000Z", "max_issues_repo_issues_event_min_datetime": "2020-10-05T10:11:24.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "vatsalg29/btp", "max_issues_repo_path": "tests/models/test_cnn_lstm.py", "max_line_length": 86, "max_stars_count": 35, "max_stars_repo_head_hexsha": "2ad4c4cfe121e951db12efa3fe5b6f1cc0188da7", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "vatsalg29/btp", "max_stars_repo_path": "tests/models/test_cnn_lstm.py", "max_stars_repo_stars_event_max_datetime": "2021-07-30T15:12:00.000Z", "max_stars_repo_stars_event_min_datetime": "2020-03-06T13:05:17.000Z", "num_tokens": 784, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2606 }
% Ubah judul dan label berikut sesuai dengan yang diinginkan. \section{Design and Implementation} \label{sec:designandimplementation} % Ubah paragraf-paragraf pada bagian ini sesuai dengan yang diinginkan. This research is explaining about the implementation of one of the branch of deep learning studies with the aim to automatically detect Indonesian hoax news by leveraging BERT method. This detection method is trained by using a combination of dataset from \url{https://data.mendeley.com/datasets/p3hfgr5j3m/1} and dataset that we made ourself for this paper alone by using web crawling technology. Picture \ref{fig:metodologi} is the outline of this research in a nutshell. \begin{figure} [h!] \centering \includegraphics[width=0.3\columnwidth]{gambar/Metodologi_Vertical_en.png} \caption{This research method in a nutshell.} \label{fig:metodologi} \end{figure} \subsection{Material and Tools Specification} The dataset that is being used in this research is a dataset originated from \url{https://data.mendeley.com/datasets/p3hfgr5j3m/1} coupled with our own made dataset in which we have create it using web crawling technology. Both of these dataset combined, is resulting in total of 1621 data with the exact details can be seen at table \ref{tab:dataset}. Meanwhile, table \ref{tab:dataset_mendeley} is the starting point of our dataset which we gotten from \url{https://data.mendeley.com/datasets/p3hfgr5j3m/1} alone. Each of these dataset is containing the content of the news along with its label which can be either "Valid" or "Hoaks". We took the news from accredited and verified news sources for the valid news, while on the other side, we took all of the hoax news mostly from \url{https://turnbackhoax.id}, a website that contains the list of user reported hoax news from many sources. \begin{table}[h] \caption{Total of news from \url{data.mendeley.com}} \label{tab:dataset_mendeley} \centering \begin{tabular}{ | l | l | } \hline \textbf{Label} & \textbf{Total Data} \\ \hline \textit{Hoaks} & 228 \\ \hline \textit{Valid} & 372 \\ \hline \textbf{Total} & \textbf{600} \\ \hline \end{tabular} \end{table} \begin{table}[h] \caption{Total of training dataset} \label{tab:dataset} \centering \begin{tabular}{ | l | l | } \hline \textbf{Label} & \textbf{Total Data} \\ \hline \textit{Hoaks} & 885 \\ \hline \textit{Valid} & 736 \\ \hline \textbf{Total} & \textbf{1621} \\ \hline \end{tabular} \end{table} \begin{table}[h] \caption{Dataset Sample} \label{tab:contoh_dataset} \centering \begin{tabular}{ | p{.8\linewidth} | l | } \hline \textbf{news} & \textbf{tagging} \\ \hline Wakil Gubernur DKI Jakarta Sandiaga Uno menargetkan pengerjaan tahap awal Stadion BMW dilakukan pada Oktober. Stadion ini diperuntukkan bagi klub Persija.... & Valid \\ \hline "Komisi II bersama KPU dan Bawaslu masih membahas ketentuan wajib cuti bagi petahana presiden yang maju Pilpres 2019. Mekanisme pengambilan..... & Valid \\ \hline Jaksa penuntut Ulumum (JPU) pada Komisi Pemberantasan Korupsi (KPK) mencecar Pejabat Pembuat Komitmen (PPK) reguler pada Direktorat Perlindungan Sosial Korban Bencana Sosial Kemensos Victorious Saut Hamonangan Siahaan soal... & Valid \\ \hline “Halo Kak! Aku Winda Dari Team Giveaway BAIM WONG Anda Memenangkan Hadiah Uang 100Jt dari kami info klik: https://wa.me/+6285796306857” & Hoax \\ \hline “Apa yang terjadi dengan hewan dalam penelitian?   Teknologi ini telah dicoba pada hewan, dan pada hewan penelitian yang dilakukan, semua hewan mati , tidak langsung dari suntikan... & Hoax \\ \hline “Kadrun istilah dr PKI alias KOMUNIS ditujukan buat islam. Kl mau jd komunis pake aja istilah kadrun buat umat islam. Auto lsg Komunis” & Hoax \\ \hline \end{tabular} \end{table} \subsection{Data Acquisition} Because the dataset that we get from \url{https://data.mendeley.com/datasets/p3hfgr5j3m/1} feel severely lacking for our purpose because it only consist of 600 data, and because there are no web crawling which outputting its result into a convenient CSV file from Indonesian news sites, we took on our hand a task to create a webcrawling program to take news content from many Indonesian news sites, those sites included but not limited to \url{liputan6.com}, \url{detik.com}, \url{tempo.com} and others. As all of those sites is rightfully accredited and verified by the government, it is used for our valid news dataset. Our hoax news site however, only has one source from \url{turnbackhoax.id}, this is mainly because said site has quite an active forum behind it in which lots of people can report their finding of hoax text, seen and checked by lots of other people, before lastly, will be uploaded to the \url{turnbackhoax.id} site. But, the biggest factor in choosing that site compare to others is mainly because \url{turnbackhoax.id} wrote the original hoaxes text in their website, this coupled with the fact that their website has some kind of structure into it has shorten our task significantly. For this research, the webcrawling process has took news from varied dates, ranging from April 2018 as the oldest to April 2021. \begin{figure} [h!] \centering \includegraphics[width=0.35\linewidth]{gambar/webcrawl method_long_en.png} \caption{Garis besar alur program \textit{web crawl}.} \label{fig:webcrawl_method} \end{figure} Picture \ref{fig:webcrawl_method} is the outline flow of the webcrawling program. Starting with inputting raw HTML code into the program, changing said code into an easier-to-process objects, get the news text and do some post-cleaning on the text, lastly, create a .CSV file to store all of the obtained news text with the appropriate format. \begin{lstlisting}[ language=HTML, caption={Penggalan Kode Sumber HTML \url{detik.com}.}, label={lst:source_detik} ] ... <div class="detail__body itp_bodycontent_wrapper"> <div class="detail__body-text itp_bodycontent"> <strong>Jakarta</strong> - Koalisi <a href="https:// detik.com/tag/jokowi" target="_blank">Jokowi</a> sedang menyusun visi-misi jagoannya. Setelah menerima masukan dari <a href="https://detik.com/ tag/muhammadiyah" target="_blank"> Muhammadiyah</a>, ... Dan kita pun membuka diri untuk menerima masukan untuk penyempurnaan," imbuhnya.<br><br><!-- s:parallaxindetail--><div class="clearfix"></div><style> ... \end{lstlisting} Firstly, we need to determine tag or class of the HTML code for our first filter. If we look into listing \ref{lst:source_detik} as a reference, we can see \texttt{detail\_\_body\-text} class is the one that containing our desired news text. We filtered that class by inputting the class name into the appropriate parameter. More often than not, our filtering result will contain some garbage or unrelated text resulting in the need to refine it further by post-clean it after the filter process. Usually, those text is writer or editorial notes, ad, or related news links which we don't need at all. Finally, the last step is outputting all of the acquired news text as a .CSV file. There are no particular reason on the article of why we chose CSV file format compared to other famous Copy of file format aside from the CSV file format is easier to use in our training program and because it is an open-format that can be opened and edited if need be, by nearly any spreadsheet program. As the general interface and improving user experience for our webcrawling software, we use a .json format file to configure what news sources that we want to get, how much is it, and when is it. All of those configuration will be processed by the program and the program will take the news in accordance with said configuration. \subsection{Preprocessing} \begin{figure}[h!] \begin{center} \includegraphics[width= .7\linewidth]{gambar/preprocess_long_en.png} \caption{Preprocessing Method} \label{fig: metodologi_preprocessing} \end{center} \end{figure} In this particular process, data need to be prepared first thing before being processed into BERT. This data preprocessing method is consisted of removing any punctuation in the text, change all of the capital letter into its lower letter and truncate any texs that is longer than the capacitites of word that BERT could process at once that is at 512 words or token. There are a few options on how to truncate the text, for example we can take the first 512 words and delete the rest, we can take the last 512 words, or we can combine from both the start of the text and the end portion of the text with some ratio. Last step of this preprocessing is to add a special \texttt{[CLS]} token. Picture \ref{fig: metodologi_preprocessing} will explain the same thing but with a better clarity. Other than that, we will also divide the dataset into 3 portionss with details stated below : \begin{itemize} \item 70\% Training, 10\% Validation, 20\% Test \end{itemize} \begin{enumerate} \item Training This set is used with BERT as an input when it is in its training phase so we can get an optimized model for our task. \item Validation This set is used right after BERT finished its training phase. Used to determined whether our created model has appropriate weight for our task or if our model still need to be trained again. This portion is also used to determine whether our model is overfitting or underfitting which is a bad thing. \item Test This set is used as an accuracy test after both the validation and the training phase is finished. The resulting accuracy of this set is the one that we consider as our result. \end{enumerate} To make it clearer, check table \ref{tab:dataset_section}. We can see based on this table that the divition of the dataset is already appropriate. \begin{table}[h] \caption{Dataset Portioning Details} \label{tab:dataset_section} \centering \begin{tabular}{ | l | l | l | l | } \hline \textbf{Bagian} & \textbf{Hoaks} & \textbf{Valid} & \textbf{Total Data} \\ \hline \textit{Training} & 647 & 519 & 1166 \\ \hline \textit{Validasi} & 85 & 78 & 163 \\ \hline \textit{Pegujian} & 153 & 139 & 292 \\ \hline \multicolumn{3}{|l|}{\textbf{Total}} & \textbf{1621} \\ \hline \end{tabular} \end{table} \subsection{The Architecture of BERT} BERT is one of the latest machine learning implementation at this time especially for Natural Language Processing (NLP) task. It is based on the Transformer implementation that is based on a previous research by Vaswani et al. \cite{attention_is_all_you_need}. BERT has successfully achieved a higher accuracy than ever before in understanding the context of a raw text if compared to other transformer implementation. One of the distinct feature of BERT is in the way it is pre-trained. There are 2 steps for pretraining BERT. The first is by doing a Masked Language Model (MLM) in which BERT will be given masked text A and some words B that can be the correct word for the masked text or not as an input, and it will need to predict whether the word B is the correct word for the masked part in text A. This way, BERT will be able to "learn" the relationship between words. The next steps for pretraining BERT is to do some Next Sentence Prediction (NSP) task. The inputted text of this task is 2 sentence, sentence A and sentence B and BERT task is to predict whether these 2 sentences will form a complete paragraph or not. By doing NSP tasks, BERT should be able to get the relationship between sentences easier. \begin{figure*}[h!] \begin{center} \includegraphics[width= 0.9\textwidth]{gambar/bert_arch.png} \caption{The Architecture of BERT in this research} \label{fig: bert_arch} \end{center} \end{figure*} In this research, we decided to use fine-tuning approach, what this mean is that we use a pre-trained BERT model rather than create our own BERT model from scratch, however, we still need to connect the last layer of BERT into a classification layer. In this case, we chose Linear Reggression as the classification layer. For greater detail, figure \ref{fig: bert_arch} is the architecture of BERT that will be used throughout this research. \subsection{Training and Validation} \begin{figure}[h!] \begin{center} \includegraphics[width= 0.9\linewidth]{gambar/training.png} \caption{Training Method} \label{fig: metodologi_training} \end{center} \end{figure} At this stage, the raw text that will be inputted into BERT has already going through its preprocessing phase and is now going into a process called Tokenizer. Tokenizer is a process to change words in a text into token according to its word embedding that is already obtained beforehand when BERT is still in its pretraining phase. Only after all of these process has done, BERT will start its training phase based on the tokenized and preprocessed data. Not all of the output of the BERT is being used in this particular research, we only need the content of the \texttt{[CLS]} token that is filled with the pooled output of all the other tokens and layers. The content of the \texttt{[CLS]} token is then inputted into Linear Regresson method. This method is choosen as it is easy enough while still retain quite good accuracy. Figure \ref{fig: metodologi_training} is the training method in a nutshell. There are also some parameters that we can adjust only in this stage, namely batch, learning rate, and epoch. Batch is a parameter to adjust how much data is being processed at once per iteration, mind you that there are usually a few iteration per epoch. By adjusting the batch values, the higher it is, the faster the training process is but at the cost of the memory usage. Thanks> to BERT method having quite a large number of layers (718 layers, in general) it can be considered quite heavy, hence we are set the batch value at 10. Epoch is how much training and validation will take place before the training phase is considered as final. This parameter is one of the most important parameters to adjust as it has a direct effect on the accuracy and the loss of our model. If our loss is too high but the accuracy is too low, it is a clear indication that our model is suffering from underfitting state, meanwhile, if our loss is too low while the accuracy is too high we still need to check if our model is actually good or if it is suffering from overfitting. As our goals in this research is only to process text that is considered to be easier compared to processing image or video, we only set the epoch value to 10. Learning rate is how much hyperparameter is allowed to change while the model is still in training process, this in turn will change the weight of the layers while in the same process based on the feedback gotten from validation phase. We decided to use the recommended value of 0.00002 \cite{koto2020indolem} The validation process is used as a way to get the loss validation value that we can use as a comparator between the loss value that we are able to obtain from the training process and the loss validation value that we get from this process. If the loss validation value is getting higher but coupled with a loss training value that shows sign of going lower still, it is a surefire way to know that our model is suffering from overfitting. In another note, if both of our loss value in our model is quite high, then there is a high chance our model is underfitting. Both cases indicate that our model can be further optimized and requiring more training while adjusting the parameter. \subsection{Testing} After going through validation and training phase, lastly, we need to test our newly created model. Based on the result of this process, we should be able to conclude whether our model can be considered good enough for our use case, or still can be further adjusted by reconfiguring some of the parameters back at the training phase. \subsection{Performance Analysis} The last step right after testing is performance analysis on our tested model. This process is quite important to see how our model will fare in real world scenario after it is being implemented. There are a few metrics that we are using to do this process, all of these metrics is considered to be the industry standard in the world that is machine learning industry. Firstly, there are confustion matrix to categorize the prediction result based on the actual label in the dataset into 4 division. The division being True Positive (TP), False Positive (FP), True Negative (TN), and False Negative (FN). We also using Recall, Precision, and F1-Score as our performance metrics in this research.
{ "alphanum_fraction": 0.7208559745, "author": null, "avg_line_length": 88.8109452736, "converted": null, "ext": "tex", "file": null, "hexsha": "4a2cd55d7f1d7e21094829673f020774a8394012", "include": null, "lang": "TeX", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "6523992a634ac8d6fbd5397dc0c42274e00d0cc8", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "chillytaka/buku-final-eng", "max_forks_repo_path": "konten/3-arsitektur.tex", "max_issues_count": null, "max_issues_repo_head_hexsha": "6523992a634ac8d6fbd5397dc0c42274e00d0cc8", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "chillytaka/buku-final-eng", "max_issues_repo_path": "konten/3-arsitektur.tex", "max_line_length": 1338, "max_stars_count": null, "max_stars_repo_head_hexsha": "6523992a634ac8d6fbd5397dc0c42274e00d0cc8", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "chillytaka/buku-final-eng", "max_stars_repo_path": "konten/3-arsitektur.tex", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 4233, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 17851 }
import numpy as np import tensorflow as tf from tqdm import tqdm from flearn.models.client import Client from flearn.utils.tf_utils import process_grad class BaseFedarated(object): def __init__(self, params, learner, dataset): '''initialize local models, build comp. graph''' # transfer parameters to self for key, val in params.items(): setattr(self, key, val); # get number of clients users, _, _, _ = dataset num_clients = len(users) if self.clients_per_round < 1: self.clients_per_round = num_clients # create server and clients' models self.client_model = learner(*params['model_params'], self.inner_opt, self.seed) self.clients = self.setup_clients(dataset, self.client_model) print('{} Clients in Total'.format(len(self.clients))) # initialize server model params self.latest_model = self.client_model.get_params() def __del__(self): '''clean up models''' self.client_model.close() def setup_clients(self, dataset, model=None): '''instantiates clients based on given train and test data directories Return: list of Clients ''' users, groups, train_data, test_data = dataset if len(groups) == 0: groups = [None for _ in users] all_clients = [Client(u, g, train_data[u], test_data[u], model) for u, g in zip(users, groups)] return all_clients def train_error_and_loss(self): '''compute clients' train statistics Return: train stats ''' num_samples = [] tot_correct = [] losses = [] for c in self.clients: ct, cl, ns = c.train_error_and_loss() tot_correct.append(ct*1.0) num_samples.append(ns) losses.append(cl*1.0) ids = [c.id for c in self.clients] groups = [c.group for c in self.clients] return ids, groups, num_samples, tot_correct, losses def test(self): '''tests self.latest_model on given clients''' num_samples = [] tot_correct = [] self.client_model.set_params(self.latest_model) for c in self.clients: ct, ns = c.test() tot_correct.append(ct*1.0) num_samples.append(ns) ids = [c.id for c in self.clients] groups = [c.group for c in self.clients] return ids, groups, num_samples, tot_correct def save(self): pass def select_clients(self, round, num_clients=20): '''selects num_clients clients weighted by number of samples from possible_clients Args: num_clients: number of clients to select; default 20 note that within function, num_clients is set to min(num_clients, len(possible_clients)) Return: list of selected clients objects ''' num_clients = min(num_clients, len(self.clients)) np.random.seed(round) # make sure for each comparison, we are selecting the same clients each round indices = np.random.choice(range(len(self.clients)), num_clients, replace=False) return indices, np.asarray(self.clients)[indices] def aggregate(self, wsolns): '''aggregate local solutions Args: wsolns: a list of local model params Return: averaged model params ''' total_weight = 0.0 base = [0]*len(wsolns[0][1]) for (w, soln) in wsolns: # w is the number of local samples total_weight += w for i, v in enumerate(soln): base[i] += w*v.astype(np.float64) averaged_soln = [v / total_weight for v in base] return averaged_soln
{ "alphanum_fraction": 0.5793450882, "author": null, "avg_line_length": 33.6440677966, "converted": null, "ext": "py", "file": null, "hexsha": "eee456027042d54e1c310f759be8ce0727c02ce4", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2022-01-05T01:50:53.000Z", "max_forks_repo_forks_event_min_datetime": "2022-01-05T01:50:53.000Z", "max_forks_repo_head_hexsha": "4097eb447a99c7180388527a2d05974906b77eb1", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "unc-optimization/FedDR", "max_forks_repo_path": "FedDR/flearn/trainers/fedbase.py", "max_issues_count": 2, "max_issues_repo_head_hexsha": "4097eb447a99c7180388527a2d05974906b77eb1", "max_issues_repo_issues_event_max_datetime": "2022-03-27T13:37:18.000Z", "max_issues_repo_issues_event_min_datetime": "2022-03-08T12:05:22.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "unc-optimization/FedDR", "max_issues_repo_path": "FedDR/flearn/trainers/fedbase.py", "max_line_length": 109, "max_stars_count": 3, "max_stars_repo_head_hexsha": "4097eb447a99c7180388527a2d05974906b77eb1", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "unc-optimization/FedDR", "max_stars_repo_path": "FedDR/flearn/trainers/fedbase.py", "max_stars_repo_stars_event_max_datetime": "2022-03-08T21:04:43.000Z", "max_stars_repo_stars_event_min_datetime": "2022-03-01T12:00:24.000Z", "num_tokens": 839, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 3970 }
from typing import Union from pathlib import Path from glob import glob from tqdm import tqdm import json import numpy as np import torch from torch.utils.data import Dataset, DataLoader from utils.bert.utils.utils import BertTokenizer def collate_fn(data): document, section, sentence, contains_ds = data[0] section = torch.tensor(section, dtype=torch.long) sentence = torch.tensor(sentence, dtype=torch.long) contains_ds = torch.tensor(contains_ds, dtype=torch.long) return document, section, sentence, contains_ds class ShowUsDataset(Dataset): def __init__(self, dataset_path:Union[str, Path], vocab_file:Union[str, Path], max_seq:int=512): super().__init__() self.dataset_path = Path(dataset_path) self.vocab_path = Path(vocab_file) self.max_seq = max_seq self.tokenizer = BertTokenizer(str(self.vocab_path), to_lower=True) self.data, self.documents, self.sections, self.sentences = self.__load_data(self.dataset_path) def __load_data(self, dataset_path:Union[str, Path]): dataset_path = Path(dataset_path) files = [Path(f) for f in glob(str(dataset_path / '*.json'))] # (doc_idx, sec_idx, sent_idx, contains_ds) data_with_ds = [] data_without_ds = [] documents = {} sections = {} sentences = {} doc_idx, sent_idx, sec_idx = 0, 0, 0 for f_path in tqdm(files, desc='loading data...'): # document data documents[doc_idx] = f_path.stem a_data = json.load(open(f_path)) for section in a_data: # section data sections[sec_idx] = section['section_text'] for sentence in section['sentences']: # sentence data sentences[sent_idx] = sentence['text'] if len(sentence['cleaned_labels']) > 0: data_with_ds.append((doc_idx, sec_idx, sent_idx, 1)) else: data_without_ds.append((doc_idx, sec_idx, sent_idx, 0)) sent_idx += 1 sec_idx += 1 doc_idx += 1 data_with_ds = np.array(data_with_ds) data_without_ds = np.array(data_without_ds) indices = np.random.choice(range(len(data_without_ds)), size=len(data_with_ds), replace=False) data = np.concatenate([data_with_ds, data_without_ds[indices]]) np.random.shuffle(data) return data, documents, sections, sentences def __len__(self): return len(self.data) def __getitem__(self, index): # get texts doc_idx, sec_idx, sent_idx, contains_ds = self.data[index] document = self.documents[doc_idx] section = self.sections[sec_idx] sentence = self.sentences[sent_idx] section = section.replace(sentence, ' ') # text -> tokens section_tokens = self.tokenizer.tokenize(section) sentence_tokens = self.tokenizer.tokenize(sentence) # tokens -> ids section_ids = self.tokenizer.tokens2ids(section_tokens) sentence_ids = self.tokenizer.tokens2ids(sentence_tokens) return document, section_ids[:self.max_seq], sentence_ids[:self.max_seq], contains_ds
{ "alphanum_fraction": 0.6245487365, "author": null, "avg_line_length": 36.9333333333, "converted": null, "ext": "py", "file": null, "hexsha": "af03545072002df2ec1ba313b72cc498c5c3e597", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "66bf4176a9f71dc9653706ff04b29038970dea04", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "akitenkrad/show-us-the-data", "max_forks_repo_path": "datasets.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "66bf4176a9f71dc9653706ff04b29038970dea04", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "akitenkrad/show-us-the-data", "max_issues_repo_path": "datasets.py", "max_line_length": 102, "max_stars_count": null, "max_stars_repo_head_hexsha": "66bf4176a9f71dc9653706ff04b29038970dea04", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "akitenkrad/show-us-the-data", "max_stars_repo_path": "datasets.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 721, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 3324 }
from flask import Flask,request, url_for, redirect, render_template, jsonify from pycaret.regression import * import pandas as pd import pickle import numpy as np app = Flask(__name__) model = load_model('insurance_14052020') cols = ['age', 'sex', 'bmi', 'children', 'smoker', 'region'] @app.route('/') def home(): return render_template("home.html") @app.route('/predict',methods=['POST']) def predict(): int_features = [x for x in request.form.values()] final = np.array(int_features) data_unseen = pd.DataFrame([final], columns = cols) prediction = predict_model(model, data=data_unseen, round = 0) prediction = int(prediction.Label[0]) return render_template('home.html',pred='Expected Bill will be {}'.format(prediction)) @app.route('/predict_api',methods=['POST']) def predict_api(): data = request.get_json(force=True) data_unseen = pd.DataFrame([data]) prediction = predict_model(model, data=data_unseen) output = prediction.Label[0] return jsonify(output) if __name__ == '__main__': app.run(host='0.0.0.0', debug=False)
{ "alphanum_fraction": 0.7019319227, "author": null, "avg_line_length": 31.9705882353, "converted": null, "ext": "py", "file": null, "hexsha": "6b2df42d48ced47ae8de4538d43436a05825ac8f", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2020-08-24T12:17:11.000Z", "max_forks_repo_forks_event_min_datetime": "2020-08-24T12:17:11.000Z", "max_forks_repo_head_hexsha": "f994f96cf310d77c913ee9ccc705f0a92ef87862", "max_forks_repo_licenses": [ "RSA-MD" ], "max_forks_repo_name": "hhaydar/ml-app-insurance", "max_forks_repo_path": "app.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "f994f96cf310d77c913ee9ccc705f0a92ef87862", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "RSA-MD" ], "max_issues_repo_name": "hhaydar/ml-app-insurance", "max_issues_repo_path": "app.py", "max_line_length": 90, "max_stars_count": null, "max_stars_repo_head_hexsha": "f994f96cf310d77c913ee9ccc705f0a92ef87862", "max_stars_repo_licenses": [ "RSA-MD" ], "max_stars_repo_name": "hhaydar/ml-app-insurance", "max_stars_repo_path": "app.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 264, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1087 }
import random import numpy as np from rl.policies import Policy class RandomPolicy(Policy): def __init__(self, action_mapper, epsilon=0.1, epsilon_decay_step=0.0001, epsilon_decay_epoch=0., epsilon_min=0.001, reset_on_game_changed=True, dropout=0.4, reset_on_epoch_start=False): super().__init__(action_mapper) self.current_epsilon = epsilon self._initial_epsilon = epsilon self._epsilon_decay_step = epsilon_decay_step self._epsilon_decay_epoch = epsilon_decay_epoch self._epsilon_min = epsilon_min self._reset_on_game_changed = reset_on_game_changed self._reset_on_epoch_start = reset_on_epoch_start if reset_on_epoch_start: assert epsilon_decay_epoch == 0, "Decay per level should be 0 if epsilon is reset each epoch." self._dropout = dropout self._is_dropout = False def allows_async_training(self): return False def epoch_start(self): dropout_chance = random.uniform(0., 1.) self._is_dropout = False if dropout_chance < self._dropout: self._is_dropout = True elif self._reset_on_epoch_start: self.current_epsilon = self._initial_epsilon else: self.current_epsilon = max( self._epsilon_min, self.current_epsilon - self._epsilon_decay_epoch ) def game_changed(self): if self._reset_on_game_changed: self.current_epsilon = self._initial_epsilon def get_action(self, env, qvalues): turn_epsilon = random.uniform(0., 1.) if not self._is_dropout and turn_epsilon < self.current_epsilon: action = env.action_space.sample() else: action = super().get_action(env, qvalues) if not self._is_dropout: self.current_epsilon = max(self._epsilon_min, self.current_epsilon - self._epsilon_decay_step) return action
{ "alphanum_fraction": 0.618556701, "author": null, "avg_line_length": 32.8307692308, "converted": null, "ext": "py", "file": null, "hexsha": "de1d659f25a2a07037386aab08679f19ddc5949d", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "a8ffdafa70d735e609296b13b0aa9950f73cfb07", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "valiro21/MarioLearningCompany", "max_forks_repo_path": "rl/policies/RandomPolicy.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "a8ffdafa70d735e609296b13b0aa9950f73cfb07", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "valiro21/MarioLearningCompany", "max_issues_repo_path": "rl/policies/RandomPolicy.py", "max_line_length": 106, "max_stars_count": null, "max_stars_repo_head_hexsha": "a8ffdafa70d735e609296b13b0aa9950f73cfb07", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "valiro21/MarioLearningCompany", "max_stars_repo_path": "rl/policies/RandomPolicy.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 434, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2134 }
#!/usr/bin/env python # A simple script to test the installed version of numpy by calling # 'numpy.test()'. Key features: # -- convenient command-line syntax # -- sets exit status appropriately, useful for automated test environments # It would be better to set this up as a module in the numpy namespace, so # that it could be run as: # python -m numpy.run_tests <args> # But, python2.4's -m switch only works with top-level modules, not modules # that are inside packages. So, once we drop 2.4 support, maybe... import sys # In case we are run from the source directory, we don't want to import numpy # from there, we want to import the installed version: sys.path.pop(0) from optparse import OptionParser parser = OptionParser("usage: %prog [options] -- [nosetests options]") parser.add_option("-v", "--verbose", action="count", dest="verbose", default=1, help="increase verbosity") parser.add_option("--doctests", action="store_true", dest="doctests", default=False, help="Run doctests in module") parser.add_option("--coverage", action="store_true", dest="coverage", default=False, help="report coverage of NumPy code (requires 'coverage' module") parser.add_option("-m", "--mode", action="store", dest="mode", default="fast", help="'fast', 'full', or something that could be " "passed to nosetests -A [default: %default]") (options, args) = parser.parse_args() import numpy result = numpy.test(options.mode, verbose=options.verbose, extra_argv=args, doctests=options.doctests, coverage=options.coverage) if result.wasSuccessful(): sys.exit(0) else: sys.exit(1)
{ "alphanum_fraction": 0.63304253, "author": null, "avg_line_length": 39.0212765957, "converted": null, "ext": "py", "file": null, "hexsha": "91e619e96e1e9e609ed76b6706976d86df264724", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 3, "max_forks_repo_forks_event_max_datetime": "2021-07-06T21:12:50.000Z", "max_forks_repo_forks_event_min_datetime": "2020-06-08T05:14:16.000Z", "max_forks_repo_head_hexsha": "665a00aec896344cfb12599add600f27a5f519d3", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "ewmoore/numpy", "max_forks_repo_path": "tools/test-installed-numpy.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "665a00aec896344cfb12599add600f27a5f519d3", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "ewmoore/numpy", "max_issues_repo_path": "tools/test-installed-numpy.py", "max_line_length": 83, "max_stars_count": 5, "max_stars_repo_head_hexsha": "665a00aec896344cfb12599add600f27a5f519d3", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "ewmoore/numpy", "max_stars_repo_path": "tools/test-installed-numpy.py", "max_stars_repo_stars_event_max_datetime": "2022-01-11T00:36:48.000Z", "max_stars_repo_stars_event_min_datetime": "2019-10-02T13:32:41.000Z", "num_tokens": 406, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1834 }
from __future__ import absolute_import from __future__ import print_function from __future__ import division import tensorflow as tf # import numpy as np import time import discogan import sys sys.path.insert(0, '../') import image_utils as iu from datasets import Pix2PixDataSet as DataSets results = { 'sample_output': './gen_img/', 'model': './model/DiscoGAN-model.ckpt' } paras = { 'epoch': 200, 'batch_size': 64, 'logging_interval': 5 } def main(): start_time = time.time() # clocking start # Dataset dataset = DataSets(height=64, width=64, channel=3, ds_path='D:/DataSets/pix2pix/', ds_name="vangogh2photo") config = tf.ConfigProto() config.gpu_options.allow_growth = True with tf.Session(config=config) as s: # DiscoGAN model model = discogan.DiscoGAN(s) # load model & graph & weight global_step = 0 ckpt = tf.train.get_checkpoint_state('./model/') if ckpt and ckpt.model_checkpoint_path: # Restores from checkpoint model.saver.restore(s, ckpt.model_checkpoint_path) global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1] print("[+] global step : %s" % global_step, " successfully loaded") else: print('[-] No checkpoint file found') # initializing variables tf.global_variables_initializer().run() d_overpowered = False # G loss > D loss * 2 for epoch in range(paras['epoch']): for step in range(1000): offset_a = (step * paras['batch_size']) % (dataset.images_a.shape[0] - paras['batch_size']) offset_b = (step * paras['batch_size']) % (dataset.images_b.shape[0] - paras['batch_size']) # batch data set batch_a = dataset.images_a[offset_a:(offset_a + paras['batch_size']), :] batch_b = dataset.images_b[offset_b:(offset_b + paras['batch_size']), :] # update D network if not d_overpowered: s.run(model.d_op, feed_dict={model.A: batch_a}) # update G network s.run(model.g_op, feed_dict={model.B: batch_b}) if epoch % paras['logging_interval'] == 0: d_loss, g_loss, summary = s.run([ model.d_loss, model.g_loss, model.merged ], feed_dict={ model.A: batch_a, model.B: batch_b }) # print loss print("[+] Epoch %03d Step %04d => " % (epoch, global_step), " D loss : {:.8f}".format(d_loss), " G loss : {:.8f}".format(g_loss)) # update overpowered d_overpowered = d_loss < g_loss / 2. # training G model with sample image and noise ab_samples = s.run(model.G_s2b, feed_dict={model.A: batch_a}) ba_samples = s.run(model.G_b2s, feed_dict={model.B: batch_b}) # summary saver model.writer.add_summary(summary, global_step=global_step) # export image generated by model G sample_image_height = model.sample_size sample_image_width = model.sample_size sample_ab_dir = results['sample_output'] + 'train_A_{0}_{1}.png'.format(epoch, global_step) sample_ba_dir = results['sample_output'] + 'train_B_{0}_{1}.png'.format(epoch, global_step) # Generated image save iu.save_images(ab_samples, size=[sample_image_height, sample_image_width], image_path=sample_ab_dir) iu.save_images(ba_samples, size=[sample_image_height, sample_image_width], image_path=sample_ba_dir) # model save model.saver.save(s, results['model'], global_step=global_step) end_time = time.time() - start_time # elapsed time print("[+] Elapsed time {:.8f}s".format(end_time)) # close tf.Session s.close() if __name__ == '__main__': main()
{ "alphanum_fraction": 0.5371993706, "author": null, "avg_line_length": 34.7578125, "converted": null, "ext": "py", "file": null, "hexsha": "83a84321ea0a0bc9570475d0ce3c63e9712bd0ca", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-08-16T01:35:21.000Z", "max_forks_repo_forks_event_min_datetime": "2021-08-16T01:35:21.000Z", "max_forks_repo_head_hexsha": "6e20e9cd07d0ec413a187d496159b97d793dab0c", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "Psyche-mia/Awesome-GANs", "max_forks_repo_path": "DiscoGAN/discogan_train.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "6e20e9cd07d0ec413a187d496159b97d793dab0c", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "Psyche-mia/Awesome-GANs", "max_issues_repo_path": "DiscoGAN/discogan_train.py", "max_line_length": 111, "max_stars_count": 1, "max_stars_repo_head_hexsha": "6e20e9cd07d0ec413a187d496159b97d793dab0c", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "sumersumerdjl/kozistr-Awesome-GANs", "max_stars_repo_path": "DiscoGAN/discogan_train.py", "max_stars_repo_stars_event_max_datetime": "2021-08-16T01:40:46.000Z", "max_stars_repo_stars_event_min_datetime": "2021-08-16T01:40:46.000Z", "num_tokens": 957, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4449 }
//============================================================================================================= /** * @file mne_project_to_surface.cpp * @author Jana Kiesel <jana.kiesel@tu-ilmenau.de>; * Matti Hamalainen <msh@nmr.mgh.harvard.edu> * @version 1.0 * @date August, 2016 * * @section LICENSE * * Copyright (C) 2016, Jana Kiesel and Matti Hamalainen. All rights reserved. * * Redistribution and use in source and binary forms, with or without modification, are permitted provided that * the following conditions are met: * * Redistributions of source code must retain the above copyright notice, this list of conditions and the * following disclaimer. * * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and * the following disclaimer in the documentation and/or other materials provided with the distribution. * * Neither the name of MNE-CPP authors nor the names of its contributors may be used * to endorse or promote products derived from this software without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A * PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. * * * @brief MNEProjectToSurface class definition. * */ //************************************************************************************************************* //============================================================================================================= // INCLUDES //============================================================================================================= #include "mne_project_to_surface.h" //************************************************************************************************************* //============================================================================================================= // INCLUDES //============================================================================================================= #include <mne/mne_bem_surface.h> #include <mne/mne_surface.h> //************************************************************************************************************* //============================================================================================================= // QT INCLUDES //============================================================================================================= //************************************************************************************************************* //============================================================================================================= // Eigen INCLUDES //============================================================================================================= #include <Eigen/Geometry> //************************************************************************************************************* //============================================================================================================= // USED NAMESPACES //============================================================================================================= using namespace MNELIB; using namespace Eigen; //************************************************************************************************************* //============================================================================================================= // DEFINE GLOBAL METHODS //============================================================================================================= //************************************************************************************************************* //============================================================================================================= // DEFINE MEMBER METHODS //============================================================================================================= MNEProjectToSurface::MNEProjectToSurface() : r1(MatrixX3f::Zero(1,3)) , r12(MatrixX3f::Zero(1,3)) , r13(MatrixX3f::Zero(1,3)) , nn(MatrixX3f::Zero(1,3)) , a(VectorXf::Zero(1)) , b(VectorXf::Zero(1)) , c(VectorXf::Zero(1)) , det(VectorXf::Zero(1)) { } //************************************************************************************************************* MNEProjectToSurface::MNEProjectToSurface(const MNEBemSurface &p_MNEBemSurf) : r1(MatrixX3f::Zero(p_MNEBemSurf.ntri,3)) , r12(MatrixX3f::Zero(p_MNEBemSurf.ntri,3)) , r13(MatrixX3f::Zero(p_MNEBemSurf.ntri,3)) , nn(MatrixX3f::Zero(p_MNEBemSurf.ntri,3)) , a(VectorXf::Zero(p_MNEBemSurf.ntri)) , b(VectorXf::Zero(p_MNEBemSurf.ntri)) , c(VectorXf::Zero(p_MNEBemSurf.ntri)) , det(VectorXf::Zero(p_MNEBemSurf.ntri)) { for (int i = 0; i < p_MNEBemSurf.ntri; ++i) { r1.row(i) = p_MNEBemSurf.rr.row(p_MNEBemSurf.tris(i,0)); r12.row(i) = p_MNEBemSurf.rr.row(p_MNEBemSurf.tris(i,1)) - r1.row(i); r13.row(i) = p_MNEBemSurf.rr.row(p_MNEBemSurf.tris(i,2)) - r1.row(i); a(i) = r12.row(i) * r12.row(i).transpose(); b(i) = r13.row(i) * r13.row(i).transpose(); c(i) = r12.row(i) * r13.row(i).transpose(); } if (!(p_MNEBemSurf.tri_nn.isZero(0))) { nn = p_MNEBemSurf.tri_nn.cast<float>(); } else { for (int i = 0; i < p_MNEBemSurf.ntri; ++i) { nn.row(i) = r12.row(i).transpose().cross(r13.row(i).transpose()).transpose(); } } det = (a.array()*b.array() - c.array()*c.array()).matrix(); } //************************************************************************************************************* MNEProjectToSurface::MNEProjectToSurface(const MNESurface &p_MNESurf) : r1(MatrixX3f::Zero(p_MNESurf.ntri,3)) , r12(MatrixX3f::Zero(p_MNESurf.ntri,3)) , r13(MatrixX3f::Zero(p_MNESurf.ntri,3)) , nn(MatrixX3f::Zero(p_MNESurf.ntri,3)) , a(VectorXf::Zero(p_MNESurf.ntri)) , b(VectorXf::Zero(p_MNESurf.ntri)) , c(VectorXf::Zero(p_MNESurf.ntri)) , det(VectorXf::Zero(p_MNESurf.ntri)) { for (int i = 0; i < p_MNESurf.ntri; ++i) { r1.row(i) = p_MNESurf.rr.row(p_MNESurf.tris(i,0)); r12.row(i) = p_MNESurf.rr.row(p_MNESurf.tris(i,1)) - r1.row(i); r13.row(i) = p_MNESurf.rr.row(p_MNESurf.tris(i,2)) - r1.row(i); nn.row(i) = r12.row(i).transpose().cross(r13.row(i).transpose()).transpose(); a(i) = r12.row(i) * r12.row(i).transpose(); b(i) = r13.row(i) * r13.row(i).transpose(); c(i) = r12.row(i) * r13.row(i).transpose(); } det = (a.array()*b.array() - c.array()*c.array()).matrix(); } //************************************************************************************************************* bool MNEProjectToSurface::mne_find_closest_on_surface(const MatrixXf &r, const int np, MatrixXf &rTri, VectorXi &nearest, VectorXf &dist) { nearest.resize(np); dist.resize(np); if (this->r1.isZero(0)) { qDebug() << "No surface loaded to make the projection./n"; return false; } int bestTri = -1; float bestDist = -1; Vector3f rTriK; for (int k = 0; k < np; ++k) { /* * To do: decide_search_restriction for the use in an iterative closest point to plane algorithm * For now it's OK to go through all triangles. */ if (!this->mne_project_to_surface(r.row(k).transpose(), rTriK, bestTri, bestDist)) { qDebug() << "The projection of point number " << k << " didn't work./n"; return false; } rTri.row(k) = rTriK.transpose(); nearest[k] = bestTri; dist[k] = bestDist; } return true; } //************************************************************************************************************* bool MNEProjectToSurface::mne_project_to_surface(const Vector3f &r, Vector3f &rTri, int &bestTri, float &bestDist) { float p = 0, q = 0, p0 = 0, q0 = 0, dist0 = 0; bestDist = 0.0f; bestTri = -1; for (int tri = 0; tri < a .size(); ++tri) { if (!this->nearest_triangle_point(r, tri, p0, q0, dist0)) { qDebug() << "The projection on triangle " << tri << " didn't work./n"; return false; } if ((bestTri < 0) || (std::fabs(dist0) < std::fabs(bestDist))) { bestDist = dist0; p = p0; q = q0; bestTri = tri; } } if (bestTri >= 0) { if (!this->project_to_triangle(rTri, p, q, bestTri)) { qDebug() << "The coordinate transform to cartesian system didn't work./n"; return false; } return true; } qDebug() << "No best Triangle found./n"; return false; } //************************************************************************************************************* bool MNEProjectToSurface::nearest_triangle_point(const Vector3f &r, const int tri, float &p, float &q, float &dist) { //Calculate some helpers Vector3f rr = r - this->r1.row(tri).transpose(); //Vector from triangle corner #1 to r float v1 = this->r12.row(tri)*rr; float v2 = this->r13.row(tri)*rr; //Calculate the orthogonal projection of the point r on the plane dist = this->nn.row(tri)*rr; p = (this->b(tri)*v1 - this->c(tri)*v2)/det(tri); q = (this->a(tri)*v2 - this->c(tri)*v1)/det(tri); //If the point projects into the triangle we are done if (p >= 0.0 && p <= 1.0 && q >= 0.0 && q <= 1.0 && (p+q) <= 1.0) { return true; } /* * Tough: must investigate the sides * We might do something intelligent here. However, for now it is ok * to do it in the hard way */ float p0, q0, t0, dist0, best, bestp, bestq; /* * Side 1 -> 2 */ p0 = p + (q * this->c(tri)) / this->a(tri); // Place the point in the corner if it is not on the side if (p0 < 0.0) { p0 = 0.0; } else if (p0 > 1.0) { p0 = 1.0; } q0 = 0; // Distance dist0 = sqrt((p-p0)*(p-p0)*this->a(tri) + (q-q0)*(q-q0)*this->b(tri) + 2*(p-p0)*(q-q0)*this->c(tri) + dist*dist); best = dist0; bestp = p0; bestq = q0; /* * Side 2 -> 3 */ t0 = ((a(tri)-c(tri))*(-p) + (b(tri)-c(tri))*q)/(a(tri)+b(tri)-2*c(tri)); // Place the point in the corner if it is not on the side if (t0 < 0.0) { t0 = 0.0; } else if (t0 > 1.0) { t0 = 1.0; } p0 = 1.0 - t0; q0 = t0; // Distance dist0 = sqrt((p-p0)*(p-p0)*this->a(tri) + (q-q0)*(q-q0)*this->b(tri) + 2*(p-p0)*(q-q0)*this->c(tri) + dist*dist); if (dist0 < best) { best = dist0; bestp = p0; bestq = q0; } /* * Side 1 -> 3 */ p0 = 0.0; q0 = q + (p * c(tri))/b(tri); // Place the point in the corner if it is not on the side if (q0 < 0.0) { q0 = 0.0; } else if (q0 > 1.0) { q0 = 1.0; } // Distance dist0 = sqrt((p-p0)*(p-p0)*this->a(tri) + (q-q0)*(q-q0)*this->b(tri) + 2*(p-p0)*(q-q0)*this->c(tri) + dist*dist); if (dist0 < best) { best = dist0; bestp = p0; bestq = q0; } dist = best; p = bestp; q = bestq; return true; } //************************************************************************************************************* bool MNEProjectToSurface::project_to_triangle(Vector3f &rTri, const float p, const float q, const int tri) { rTri = this->r1.row(tri) + p*this->r12.row(tri) + q*this->r13.row(tri); return true; }
{ "alphanum_fraction": 0.4358686049, "author": null, "avg_line_length": 35.0732394366, "converted": null, "ext": "cpp", "file": null, "hexsha": "84713ba83b656e77f04261c025ef40951c1d3770", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "36f21b3ab0c65a133027da83fa8e2a652acd1485", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "ChunmingGu/mne-cpp-master", "max_forks_repo_path": "libraries/mne/mne_project_to_surface.cpp", "max_issues_count": 1, "max_issues_repo_head_hexsha": "36f21b3ab0c65a133027da83fa8e2a652acd1485", "max_issues_repo_issues_event_max_datetime": "2018-08-23T12:40:56.000Z", "max_issues_repo_issues_event_min_datetime": "2018-08-23T12:40:56.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "ChunmingGu/mne-cpp-master", "max_issues_repo_path": "libraries/mne/mne_project_to_surface.cpp", "max_line_length": 116, "max_stars_count": 1, "max_stars_repo_head_hexsha": "36f21b3ab0c65a133027da83fa8e2a652acd1485", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "ChunmingGu/mne-cpp-master", "max_stars_repo_path": "libraries/mne/mne_project_to_surface.cpp", "max_stars_repo_stars_event_max_datetime": "2019-05-14T07:38:25.000Z", "max_stars_repo_stars_event_min_datetime": "2019-05-14T07:38:25.000Z", "num_tokens": 3081, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 12451 }
#include <boost/python/class.hpp> #include <boost/python/overloads.hpp> #include <boost/python/manage_new_object.hpp> #include <boost/python/suite/indexing/vector_indexing_suite.hpp> #include "MultiChannel.h" using namespace boost::python; // // MultiChannel class // void wrapMultiChannel() { #if PVA_API_VERSION >= 481 class_<MultiChannel>("MultiChannel", "This class is used to communicate with multiple PV channels.\n\n" "**MultiChannel(names [, providerType=PVA])**\n\n" "\t:Parameter: *names* (list) - channel names\n\n" "\t:Parameter: *providerType* (PROVIDERTYPE) - provider type, either PVA (PV Access) or CA (Channel Access)\n\n" "\tThe following example allows access to PVA channels 'int01' and 'double01':\n\n" "\t::\n\n" "\t\tmChannel = MultiChannel(['int01','double01'])\n\n", init<const boost::python::list&>()) .def(init<const boost::python::list&, PvProvider::ProviderType>()) // // Get methods // .def("get", static_cast<PvObject*(MultiChannel::*)(const std::string&)>(&MultiChannel::get), return_value_policy<manage_new_object>(), args("requestDescriptor"), "Retrieves PV data from multiple channels.\n\n" ":Parameter: *requestDescriptor* (str) - PV request descriptor\n\n" ":Returns: PvObject with NTMultiChannel structure that contains retrieved data from all member channels as a variant union array\n\n" "::\n\n" " pv = mChannel.get('field(value,alarm)')\n\n") .def("get", static_cast<PvObject*(MultiChannel::*)()>(&MultiChannel::get), return_value_policy<manage_new_object>(), "Retrieves PV data from multiple channels using the default request descriptor 'field(value,alarm,timeStamp)'.\n\n" ":Returns: PvObject with NTMultiChannel structure that contains retrieved data from all member channels as a variant union array\n\n" "::\n\n" " pv = mChannel.get()\n\n") .def("getAsDoubleArray", static_cast<boost::python::list(MultiChannel::*)()>(&MultiChannel::getAsDoubleArray), "Retrieves PV data from multiple channels as array of doubles.\n\n" ":Returns: list of floats\n\n" "::\n\n" " valueList = mChannel.getAsDoubleArray()\n\n") .def("put", static_cast<void(MultiChannel::*)(const boost::python::list&)>(&MultiChannel::put), args("pvObjectList"), "Assigns data to multi-channel member PVs'.\n\n" ":Parameter: *pvObjectList* (list) - list of PvObject instances that will be assigned to the multi-channel PVs\n\n" "::\n\n" " mChannel = MultiChannel(['PVRstringArray', 'PVRdouble'])\n\n" " pv1 = PvObject({'value' : [STRING]}, {'value' : ['ccc', 'ddd']})\n\n" " pv2 = PvDouble(44.44)\n\n" " mChannel.put([pv1,pv2])\n\n") .def("putAsDoubleArray", static_cast<void(MultiChannel::*)(const boost::python::list&)>(&MultiChannel::putAsDoubleArray), args("valueList"), "Assigns data to multi-channel member PVs'.\n\n" ":Parameter: *valueList* (list) - list of float values that will be assigned to the multi-channel PVs\n\n" "::\n\n" " mChannel = MultiChannel(['PVRdouble01', 'PVRdouble02'])\n\n" " mChannel.put([1.0,2.0])\n\n") .def("monitor", static_cast<void(MultiChannel::*)(const boost::python::object&)>(&MultiChannel::monitor), args("subscriber"), "Starts multi-channel monitor with default poll period and request descriptor 'field(value,alarm,timeStamp)'.\n\n" ":Parameter: *subscriber* (object) - reference to python function that will be executed when PV value changes; the function should take PvObject instance as its argument\n\n" "::\n\n" " def echo(pvObject):\n\n" " print('New PV values: %s' % pvObject)\n\n" " mChannel.monitor(echo)\n\n") .def("monitor", static_cast<void(MultiChannel::*)(const boost::python::object&, double)>(&MultiChannel::monitor), args("subscriber", "pollPeriod"), "Starts multi-channel monitor with default request descriptor 'field(value,alarm,timeStamp)'.\n\n" ":Parameter: *subscriber* (object) - reference to python function that will be executed when PV value changes; the function should take PvObject instance as its argument\n\n" ":Parameter: *pollPeriod* (float) - period in seconds between two multi-channel polls\n\n" "::\n\n" " def echo(pvObject):\n\n" " print('New PV values: %s' % pvObject)\n\n" " mChannel.monitor(echo, 1.0)\n\n") .def("monitor", static_cast<void(MultiChannel::*)(const boost::python::object&, double, const std::string&)>(&MultiChannel::monitor), args("subscriber", "pollPeriod", "requestDescriptor"), "Starts multi-channel monitor.\n\n" ":Parameter: *subscriber* (object) - reference to python function that will be executed when PV value changes; the function should take PvObject instance as its argument\n\n" ":Parameter: *pollPeriod* (float) - period in seconds between two multi-channel polls\n\n" ":Parameter: *requestDescriptor* (str) - describes what PV data should be sent to subscribed channel clients\n\n" "::\n\n" " def echo(pvObject):\n\n" " print('New PV values: %s' % pvObject)\n\n" " mChannel.monitor(echo, 1.0, 'field(value,alarm,timeStamp)')\n\n") .def("monitorAsDoubleArray", static_cast<void(MultiChannel::*)(const boost::python::object&)>(&MultiChannel::monitorAsDoubleArray), args("subscriber"), "Starts multi-channel monitor for processing list of double values.\n\n" ":Parameter: *subscriber* (object) - reference to python function that will be executed when PV values change; the function should take list of python floats as its argument\n\n" "::\n\n" " def echo(valueList):\n\n" " print('New PV values: %s' % x)\n\n" " mChannel.monitorAsDoubleArray(echo, 1.0)\n\n") .def("monitorAsDoubleArray", static_cast<void(MultiChannel::*)(const boost::python::object&, double)>(&MultiChannel::monitorAsDoubleArray), args("subscriber", "pollPeriod"), "Starts multi-channel monitor for processing list of double values.\n\n" ":Parameter: *subscriber* (object) - reference to python function that will be executed when PV values change; the function should take list of python floats as its argument\n\n" ":Parameter: *pollPeriod* (float) - period in seconds between two multi-channel polls\n\n" "::\n\n" " def echo(valueList):\n\n" " print('New PV values: %s' % x)\n\n" " mChannel.monitorAsDoubleArray(echo, 1.0)\n\n") .def("stopMonitor", &MultiChannel::stopMonitor, "Stops multi-channel monitor for PV value changes.\n\n" "::\n\n" " mChannel.stopMonitor()\n\n") ; #endif // if PVA_API_VERSION >= 481 } // wrapChannel()
{ "alphanum_fraction": 0.6396510483, "author": null, "avg_line_length": 49.0137931034, "converted": null, "ext": "cpp", "file": null, "hexsha": "83744930db4cba5a0809579242be1e9a67c0e606", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 26, "max_forks_repo_forks_event_max_datetime": "2021-08-24T08:26:32.000Z", "max_forks_repo_forks_event_min_datetime": "2015-03-31T23:20:27.000Z", "max_forks_repo_head_hexsha": "71c6cf56c76221b718ebd11fefcade194fcfdd28", "max_forks_repo_licenses": [ "CNRI-Python" ], "max_forks_repo_name": "mrkraimer/pvaPy", "max_forks_repo_path": "src/pvaccess/pvaccess.MultiChannel.cpp", "max_issues_count": 63, "max_issues_repo_head_hexsha": "71c6cf56c76221b718ebd11fefcade194fcfdd28", "max_issues_repo_issues_event_max_datetime": "2022-03-10T23:14:25.000Z", "max_issues_repo_issues_event_min_datetime": "2015-06-11T14:12:55.000Z", "max_issues_repo_licenses": [ "CNRI-Python" ], "max_issues_repo_name": "mrkraimer/pvaPy", "max_issues_repo_path": "src/pvaccess/pvaccess.MultiChannel.cpp", "max_line_length": 186, "max_stars_count": 19, "max_stars_repo_head_hexsha": "71c6cf56c76221b718ebd11fefcade194fcfdd28", "max_stars_repo_licenses": [ "CNRI-Python" ], "max_stars_repo_name": "mrkraimer/pvaPy", "max_stars_repo_path": "src/pvaccess/pvaccess.MultiChannel.cpp", "max_stars_repo_stars_event_max_datetime": "2022-02-24T10:47:28.000Z", "max_stars_repo_stars_event_min_datetime": "2015-02-09T08:58:19.000Z", "num_tokens": 1812, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 7107 }
SUBROUTINE ML5_0_HELAS_CALLS_AMPB_1(P,NHEL,H,IC) C C Modules C USE ML5_0_POLYNOMIAL_CONSTANTS C IMPLICIT NONE C C CONSTANTS C INTEGER NEXTERNAL PARAMETER (NEXTERNAL=4) INTEGER NCOMB PARAMETER (NCOMB=16) INTEGER NBORNAMPS PARAMETER (NBORNAMPS=3) INTEGER NLOOPS, NLOOPGROUPS, NCTAMPS PARAMETER (NLOOPS=44, NLOOPGROUPS=28, NCTAMPS=85) INTEGER NLOOPAMPS PARAMETER (NLOOPAMPS=129) INTEGER NWAVEFUNCS,NLOOPWAVEFUNCS PARAMETER (NWAVEFUNCS=10,NLOOPWAVEFUNCS=93) REAL*8 ZERO PARAMETER (ZERO=0D0) REAL*16 MP__ZERO PARAMETER (MP__ZERO=0.0E0_16) C These are constants related to the split orders INTEGER NSO, NSQUAREDSO, NAMPSO PARAMETER (NSO=1, NSQUAREDSO=1, NAMPSO=2) C C ARGUMENTS C REAL*8 P(0:3,NEXTERNAL) INTEGER NHEL(NEXTERNAL), IC(NEXTERNAL) INTEGER H C C LOCAL VARIABLES C INTEGER I,J,K COMPLEX*16 COEFS(MAXLWFSIZE,0:VERTEXMAXCOEFS-1,MAXLWFSIZE) LOGICAL DUMMYFALSE DATA DUMMYFALSE/.FALSE./ C C GLOBAL VARIABLES C INCLUDE 'coupl.inc' INCLUDE 'mp_coupl.inc' INTEGER HELOFFSET INTEGER GOODHEL(NCOMB) LOGICAL GOODAMP(NSQUAREDSO,NLOOPGROUPS) COMMON/ML5_0_FILTERS/GOODAMP,GOODHEL,HELOFFSET LOGICAL CHECKPHASE LOGICAL HELDOUBLECHECKED COMMON/ML5_0_INIT/CHECKPHASE, HELDOUBLECHECKED INTEGER SQSO_TARGET COMMON/ML5_0_SOCHOICE/SQSO_TARGET LOGICAL UVCT_REQ_SO_DONE,MP_UVCT_REQ_SO_DONE,CT_REQ_SO_DONE $ ,MP_CT_REQ_SO_DONE,LOOP_REQ_SO_DONE,MP_LOOP_REQ_SO_DONE $ ,CTCALL_REQ_SO_DONE,FILTER_SO COMMON/ML5_0_SO_REQS/UVCT_REQ_SO_DONE,MP_UVCT_REQ_SO_DONE $ ,CT_REQ_SO_DONE,MP_CT_REQ_SO_DONE,LOOP_REQ_SO_DONE $ ,MP_LOOP_REQ_SO_DONE,CTCALL_REQ_SO_DONE,FILTER_SO INTEGER I_SO COMMON/ML5_0_I_SO/I_SO INTEGER I_LIB COMMON/ML5_0_I_LIB/I_LIB COMPLEX*16 AMP(NBORNAMPS) COMMON/ML5_0_AMPS/AMP COMPLEX*16 W(20,NWAVEFUNCS) COMMON/ML5_0_W/W COMPLEX*16 WL(MAXLWFSIZE,0:LOOPMAXCOEFS-1,MAXLWFSIZE, $ -1:NLOOPWAVEFUNCS) COMPLEX*16 PL(0:3,-1:NLOOPWAVEFUNCS) COMMON/ML5_0_WL/WL,PL COMPLEX*16 AMPL(3,NCTAMPS) COMMON/ML5_0_AMPL/AMPL C C ---------- C BEGIN CODE C ---------- C The target squared split order contribution is already reached C if true. IF (FILTER_SO.AND.CT_REQ_SO_DONE) THEN GOTO 1001 ENDIF CALL VXXXXX(P(0,1),ZERO,NHEL(1),-1*IC(1),W(1,1)) CALL VXXXXX(P(0,2),ZERO,NHEL(2),-1*IC(2),W(1,2)) CALL OXXXXX(P(0,3),MDL_MT,NHEL(3),+1*IC(3),W(1,3)) CALL IXXXXX(P(0,4),MDL_MT,NHEL(4),-1*IC(4),W(1,4)) CALL VVV1P0_1(W(1,1),W(1,2),GC_4,ZERO,ZERO,W(1,5)) C Amplitude(s) for born diagram with ID 1 CALL FFV1_0(W(1,4),W(1,3),W(1,5),GC_5,AMP(1)) CALL FFV1_1(W(1,3),W(1,1),GC_5,MDL_MT,MDL_WT,W(1,6)) C Amplitude(s) for born diagram with ID 2 CALL FFV1_0(W(1,4),W(1,6),W(1,2),GC_5,AMP(2)) CALL FFV1_2(W(1,4),W(1,1),GC_5,MDL_MT,MDL_WT,W(1,7)) C Amplitude(s) for born diagram with ID 3 CALL FFV1_0(W(1,7),W(1,3),W(1,2),GC_5,AMP(3)) CALL FFV1P0_3(W(1,4),W(1,3),GC_5,ZERO,ZERO,W(1,8)) C Counter-term amplitude(s) for loop diagram number 4 CALL R2_GG_1_R2_GG_2_0(W(1,5),W(1,8),R2_GGG_1,R2_GGG_2,AMPL(1,1)) C Counter-term amplitude(s) for loop diagram number 5 CALL FFV1_0(W(1,4),W(1,3),W(1,5),R2_GQQ,AMPL(1,2)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB_1EPS,AMPL(2,3)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB_1EPS,AMPL(2,4)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB_1EPS,AMPL(2,5)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB_1EPS,AMPL(2,6)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB_1EPS,AMPL(2,7)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB_1EPS,AMPL(2,8)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQG_1EPS,AMPL(2,9)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQB,AMPL(1,10)) CALL FFV1_0(W(1,4),W(1,3),W(1,5),UV_GQQT,AMPL(1,11)) CALL FFV1_2(W(1,4),W(1,2),GC_5,MDL_MT,MDL_WT,W(1,9)) C Counter-term amplitude(s) for loop diagram number 7 CALL R2_QQ_1_R2_QQ_2_0(W(1,9),W(1,6),R2_QQQ,R2_QQT,AMPL(1,12)) CALL R2_QQ_2_0(W(1,9),W(1,6),UV_TMASS_1EPS,AMPL(2,13)) CALL R2_QQ_2_0(W(1,9),W(1,6),UV_TMASS,AMPL(1,14)) C Counter-term amplitude(s) for loop diagram number 8 CALL FFV1_0(W(1,4),W(1,6),W(1,2),R2_GQQ,AMPL(1,15)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB_1EPS,AMPL(2,16)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB_1EPS,AMPL(2,17)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB_1EPS,AMPL(2,18)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB_1EPS,AMPL(2,19)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB_1EPS,AMPL(2,20)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB_1EPS,AMPL(2,21)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQG_1EPS,AMPL(2,22)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQB,AMPL(1,23)) CALL FFV1_0(W(1,4),W(1,6),W(1,2),UV_GQQT,AMPL(1,24)) CALL FFV1_1(W(1,3),W(1,2),GC_5,MDL_MT,MDL_WT,W(1,10)) C Counter-term amplitude(s) for loop diagram number 10 CALL R2_QQ_1_R2_QQ_2_0(W(1,7),W(1,10),R2_QQQ,R2_QQT,AMPL(1,25)) CALL R2_QQ_2_0(W(1,7),W(1,10),UV_TMASS_1EPS,AMPL(2,26)) CALL R2_QQ_2_0(W(1,7),W(1,10),UV_TMASS,AMPL(1,27)) C Counter-term amplitude(s) for loop diagram number 11 CALL FFV1_0(W(1,7),W(1,3),W(1,2),R2_GQQ,AMPL(1,28)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB_1EPS,AMPL(2,29)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB_1EPS,AMPL(2,30)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB_1EPS,AMPL(2,31)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB_1EPS,AMPL(2,32)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB_1EPS,AMPL(2,33)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB_1EPS,AMPL(2,34)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQG_1EPS,AMPL(2,35)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQB,AMPL(1,36)) CALL FFV1_0(W(1,7),W(1,3),W(1,2),UV_GQQT,AMPL(1,37)) C Counter-term amplitude(s) for loop diagram number 13 CALL FFV1_0(W(1,4),W(1,10),W(1,1),R2_GQQ,AMPL(1,38)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB_1EPS,AMPL(2,39)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB_1EPS,AMPL(2,40)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB_1EPS,AMPL(2,41)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB_1EPS,AMPL(2,42)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB_1EPS,AMPL(2,43)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB_1EPS,AMPL(2,44)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQG_1EPS,AMPL(2,45)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQB,AMPL(1,46)) CALL FFV1_0(W(1,4),W(1,10),W(1,1),UV_GQQT,AMPL(1,47)) C Counter-term amplitude(s) for loop diagram number 14 CALL FFV1_0(W(1,9),W(1,3),W(1,1),R2_GQQ,AMPL(1,48)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB_1EPS,AMPL(2,49)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB_1EPS,AMPL(2,50)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB_1EPS,AMPL(2,51)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB_1EPS,AMPL(2,52)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB_1EPS,AMPL(2,53)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB_1EPS,AMPL(2,54)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQG_1EPS,AMPL(2,55)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQB,AMPL(1,56)) CALL FFV1_0(W(1,9),W(1,3),W(1,1),UV_GQQT,AMPL(1,57)) C Counter-term amplitude(s) for loop diagram number 17 CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GG,AMPL(1,58)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB_1EPS,AMPL(2,59)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB_1EPS,AMPL(2,60)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB_1EPS,AMPL(2,61)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB_1EPS,AMPL(2,62)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB_1EPS,AMPL(2,63)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB_1EPS,AMPL(2,64)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GG_1EPS,AMPL(2,65)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GB,AMPL(1,66)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),UV_3GT,AMPL(1,67)) C Counter-term amplitude(s) for loop diagram number 31 CALL R2_GG_1_0(W(1,5),W(1,8),R2_GGQ,AMPL(1,68)) CALL R2_GG_1_0(W(1,5),W(1,8),R2_GGQ,AMPL(1,69)) CALL R2_GG_1_0(W(1,5),W(1,8),R2_GGQ,AMPL(1,70)) CALL R2_GG_1_0(W(1,5),W(1,8),R2_GGQ,AMPL(1,71)) C Counter-term amplitude(s) for loop diagram number 32 CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GQ,AMPL(1,72)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GQ,AMPL(1,73)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GQ,AMPL(1,74)) CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GQ,AMPL(1,75)) C Counter-term amplitude(s) for loop diagram number 34 CALL R2_GG_1_R2_GG_3_0(W(1,5),W(1,8),R2_GGQ,R2_GGB,AMPL(1,76)) C Counter-term amplitude(s) for loop diagram number 35 CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GQ,AMPL(1,77)) C Counter-term amplitude(s) for loop diagram number 37 CALL R2_GG_1_R2_GG_3_0(W(1,5),W(1,8),R2_GGQ,R2_GGT,AMPL(1,78)) C Counter-term amplitude(s) for loop diagram number 38 CALL VVV1_0(W(1,1),W(1,2),W(1,8),R2_3GQ,AMPL(1,79)) C At this point, all CT amps needed for (QCD=6), i.e. of split C order ID=1, are computed. IF(FILTER_SO.AND.SQSO_TARGET.EQ.1) GOTO 2000 GOTO 1001 2000 CONTINUE CT_REQ_SO_DONE=.TRUE. 1001 CONTINUE END
{ "alphanum_fraction": 0.6241134752, "author": null, "avg_line_length": 43.3348623853, "converted": null, "ext": "f", "file": null, "hexsha": "cb912048c843496f63ef640defa460288af3fb7d", "include": null, "lang": "FORTRAN", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2022-03-10T09:02:00.000Z", "max_forks_repo_forks_event_min_datetime": "2022-03-10T09:02:00.000Z", "max_forks_repo_head_hexsha": "2e04f23353051f64e1604b23105fe3faabd32869", "max_forks_repo_licenses": [ "NCSA" ], "max_forks_repo_name": "valassi/mg5amc_test", "max_forks_repo_path": "tests/input_files/IOTestsComparison/MadLoop_output_from_the_interface/TIR_output/%ggttx_IOTest%SubProcesses%P0_gg_ttx%helas_calls_ampb_1.f", "max_issues_count": 4, "max_issues_repo_head_hexsha": "2e04f23353051f64e1604b23105fe3faabd32869", "max_issues_repo_issues_event_max_datetime": "2022-03-30T16:15:01.000Z", "max_issues_repo_issues_event_min_datetime": "2022-03-10T09:13:31.000Z", "max_issues_repo_licenses": [ "NCSA" ], "max_issues_repo_name": "valassi/mg5amc_test", "max_issues_repo_path": "tests/input_files/IOTestsComparison/MadLoop_output_from_the_interface/TIR_output/%ggttx_IOTest%SubProcesses%P0_gg_ttx%helas_calls_ampb_1.f", "max_line_length": 71, "max_stars_count": null, "max_stars_repo_head_hexsha": "2e04f23353051f64e1604b23105fe3faabd32869", "max_stars_repo_licenses": [ "NCSA" ], "max_stars_repo_name": "valassi/mg5amc_test", "max_stars_repo_path": "tests/input_files/IOTestsComparison/MadLoop_output_from_the_interface/TIR_output/%ggttx_IOTest%SubProcesses%P0_gg_ttx%helas_calls_ampb_1.f", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 4656, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 9447 }
Require Import Bool List. Import ListNotations. Require Import Undecidability.PCP.PCP. Require Import Undecidability.PCP.Util.Facts. Import PCPListNotation. Require Import Undecidability.PCP.Util.PCP_facts. Require Import Undecidability.Synthetic.Definitions. Local Hint Rewrite <- app_assoc : list. Local Hint Resolve in_eq in_nil in_cons in_or_app : core. Set Default Goal Selector "!". (* * MPCP to PCP *) Section MPCP_PCP. Local Definition card := card nat. Local Definition string := string nat. Local Notation "x / y" := (x, y). Variable R : list (card). Variable x0 y0 : string. Definition Sigma := sym (x0/y0 :: R). Definition R_Sigma : sym R <<= Sigma. Proof. unfold Sigma. cbn. do 2 apply incl_appr. apply incl_refl. Qed. Definition dollar := fresh Sigma. Notation "$" := dollar. Definition hash := fresh (dollar :: Sigma). Notation "#" := hash. Fixpoint hash_l (x : string) := match x with | [] => [] | a :: x => # :: a :: hash_l x end. Notation "#_L" := hash_l. Fixpoint hash_r (x : string) := match x with | [] => [] | a :: x => a :: # :: hash_r x end. Notation "#_R" := hash_r. Definition d := ($ :: (#_L x0)) / ($ :: # :: (#_R y0)). Definition e := [#;$] / [$]. Definition P := [d;e] ++ map (fun '(x,y) => (#_L x, #_R y)) (filter (fun '(x,y) => is_cons x || is_cons y) (x0/y0::R)). Lemma P_inv c : c el P -> c = d \/ c = e \/ (exists x y, (x,y) el (x0/y0 :: R) /\ c = (#_L x, #_R y) /\ ( (x,y) <> (nil, nil))). Proof. cbn -[filter]. intros. firstorder. eapply in_map_iff in H as ((x,y) & <- & (? & ? % orb_prop) % filter_In). rewrite !is_cons_true_iff in H0. right. right. exists x, y. firstorder; destruct x, y; cbn; firstorder congruence. Qed. Lemma P_inv_top x y a : a el Sigma -> ~(a :: x,y) el P. Proof. intros ? [ | [ | (x'' & y' & ? & ? & ?) ] ] % P_inv. - inv H0. edestruct fresh_spec; eauto. - inv H0. edestruct fresh_spec with (l := dollar :: Sigma); [ | reflexivity]. firstorder. - inv H1. destruct x''; cbn -[fresh] in *; [congruence | ]. inversion H4. edestruct fresh_spec with (l := dollar :: Sigma); [ | reflexivity ]. unfold hash in *. rewrite <- H3. firstorder. Qed. Lemma P_inv_bot x y : ~(y, # :: x) el P. Proof. intros [ | [ | (x'' & y' & ? & ? & ?) ] ] % P_inv. - inv H. edestruct fresh_spec; eauto. - inv H. edestruct fresh_spec; eauto. - inv H0. destruct y'; cbn -[fresh] in *; [congruence | ]. inversion H4. edestruct fresh_spec with (l := dollar :: Sigma); [ | reflexivity ]. unfold hash in *. rewrite H2. right. eapply sym_word_R; eauto. Qed. Lemma match_start d' B : d' :: B <<= P -> tau1 (d' :: B) = tau2 (d' :: B) -> d' = d. Proof. intros Hs Hm. assert (d' el P) as [ -> | [ -> | (x & y & ? & -> & ?) ] ] % P_inv by now eapply Hs; cbn -[fresh] in Hm. - congruence. - inv Hm. now edestruct fresh_spec; eauto. - cbn in Hm. destruct x, y; try firstorder congruence. + destruct (tau1_inv Hm) as (x' & y' & ? ). pose (cons_incl Hs). assert ( (n :: x') / y' el P) as [] % P_inv_top by eauto. eapply sym_word_R; eauto. + cbn -[fresh] in Hm. symmetry in Hm. destruct (tau2_inv Hm) as (x' & y' & ? ). pose (cons_incl Hs). assert ( y' / (# :: x') el P) as [] % P_inv_bot by eauto. + cbn -[fresh] in Hm. inversion Hm. assert (fresh (dollar :: Sigma) = hash) by reflexivity. edestruct fresh_spec; try eassumption. right. eapply sym_word_R in H. subst. eauto. Qed. Lemma hash_swap x : #_L x ++ [#] = # :: #_R x. Proof. induction x; cbn in *; now try rewrite IHx. Qed. Lemma hash_L_app x y : #_L (x ++ y) = #_L x ++ #_L y. Proof. induction x; cbn in *; now try rewrite IHx. Qed. Lemma hash_R_app x y : #_R (x ++ y) = #_R x ++ #_R y. Proof. induction x; cbn in *; now try rewrite IHx. Qed. Lemma hash_L_diff x y : #_L x <> # :: #_R y. Proof. revert y. induction x; cbn -[fresh]; intros ? ?; inv H. destruct y; inv H1. firstorder. Qed. Lemma hash_R_inv x y : #_R x = #_R y -> x = y. Proof. revert y; induction x; intros [] H; inv H; firstorder congruence. Qed. Lemma doll_hash_L x: x <<= Sigma -> ~ $ el #_L x. Proof. induction x; intros. -- firstorder. -- intros [ | []]. ++ now eapply fresh_spec in H0 as []. ++ symmetry in H0. eapply fresh_spec in H0 as []. firstorder. ++ firstorder. Qed. Lemma MPCP_PCP x y A : A <<= x0/y0 :: R -> x ++ tau1 A = y ++ tau2 A -> exists B, B <<= P /\ (#_L x) ++ tau1 B = # :: #_R y ++ tau2 B. Proof. revert x y; induction A; cbn -[fresh P] in *; intros. - rewrite !app_nil_r in H0. subst. exists [e]. firstorder. cbn -[fresh]. enough ((#_L y ++ [#]) ++ [$] = # :: #_R y ++ [$]) by now autorewrite with list in *. now rewrite hash_swap. - destruct a as (x', y'). assert ( (x ++ x') ++ tau1 A = (y ++ y') ++ tau2 A) by now simpl_list. eapply IHA in H1 as (B & ? & ?) ; [ | firstorder]. rewrite hash_L_app, hash_R_app in H2. autorewrite with list in H2. destruct (is_cons x' || is_cons y') eqn:E. + exists ( (#_L x' / #_R y') :: B). split. * intros ? [ <- | ]; [ | eauto]. unfold P. rewrite in_app_iff, in_map_iff. right. exists (x', y'). split; [easy|]. eapply filter_In. eauto. * eassumption. + exists B. rewrite orb_false_iff, <- !not_true_iff_false, !is_cons_true_iff in E. destruct E. destruct x', y'; firstorder congruence. Qed. Lemma PCP_MPCP x y B : B <<= P -> x <<= Sigma -> y <<= Sigma -> (#_L x) ++ tau1 B = # :: (#_R y) ++ tau2 B -> exists A, A <<= x0/y0::R /\ x ++ tau1 A = y ++ tau2 A. Proof. revert x y. induction B; cbn -[fresh P] in *; intros. - exfalso. rewrite !app_nil_r in H2. eapply hash_L_diff in H2 as []. - assert (a el P) as [ -> | [ -> | (x' & y' & ? & -> & ?) ] ] % P_inv by intuition; cbn -[fresh P] in *. + exfalso. setoid_rewrite app_comm_cons in H2 at 2. eapply list_prefix_inv in H2 as [[] % hash_L_diff E2]. * now eapply doll_hash_L. * rewrite <- hash_swap. rewrite in_app_iff. intros [ | [ | []]]. { eapply doll_hash_L; eauto. } now eapply fresh_spec in H3 as []. + exists []. assert ((#_L x ++ [#]) ++ $ :: tau1 B = (# :: #_R y) ++ $ :: tau2 B) by now simpl_list. eapply list_prefix_inv in H3. { rewrite hash_swap in H3. inv H3. inv H4. eapply hash_R_inv in H6 as ->. firstorder. } * rewrite in_app_iff. intros [ | [ | []]]. { eapply doll_hash_L in H4; eauto. } now eapply fresh_spec in H4 as []. * rewrite <- hash_swap. rewrite in_app_iff. intros [ | [ | []]]. { eapply doll_hash_L; eauto. } now eapply fresh_spec in H4 as []. + assert ((#_L x ++ #_L x') ++ tau1 B = # :: (#_R y ++ #_R y') ++ tau2 B) by now simpl_list. rewrite <- hash_L_app, <- hash_R_app in H5. eapply IHB in H5 as (A & ? & ?). * exists (x' / y' :: A). intuition idtac; try inv H7; auto with datatypes; cbn; now autorewrite with list in *. * eapply incl_cons_inv. eassumption. * eapply incl_app. { eauto. } intros ? ?. destruct H3. { inv H3. eauto. } eapply R_Sigma, sym_word_l; eauto. * eapply incl_app. { eauto. } intros ? ?. destruct H3. { inv H3. eauto. } eapply R_Sigma, sym_word_R; eauto. Qed. Lemma MPCP_PCP_cor : MPCP (x0/y0, R) <-> PCP P. Proof. split. - intros (A & Hi & (B & HiB & H) % MPCP_PCP). + exists (d :: B). firstorder try congruence. cbn. f_equal. now rewrite H. + eassumption. - intros ([|d' B] & Hi & He & H); firstorder. pose proof H as -> % match_start; eauto. cbn -[fresh] in H. inv H. eapply PCP_MPCP in H1; cbn. + eassumption. + eapply cons_incl. eassumption. + apply incl_appl. apply incl_refl. + apply incl_appr. apply incl_appl. apply incl_refl. Qed. End MPCP_PCP. Theorem reduction : MPCP ⪯ PCP. Proof. exists (fun '(x, y, R) => P R x y). intros ((x & y) & R). eapply MPCP_PCP_cor. Qed.
{ "alphanum_fraction": null, "author": "uds-psl", "avg_line_length": null, "converted": null, "ext": null, "file": null, "hexsha": null, "include": null, "lang": null, "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": "github-repos/coq/uds-psl-coq-library-undecidability/coq-library-undecidability-4547d325e8ce7a6d841fbfe5df4429ee9cb6f214/theories/PCP/Reductions/MPCP_to_PCP.v", "reason": null, "repo": "coq-library-undecidability", "save_path": "github-repos/coq/uds-psl-coq-library-undecidability", "sha": "4547d325e8ce7a6d841fbfe5df4429ee9cb6f214", "size": null }
from tkinter import * from PIL import ImageTk, Image from tkinter import filedialog import numpy as np from tensorflow import keras import tensorflow as tf from tensorflow.python.keras.models import load_model from tkinter import messagebox longitud, altura = 200, 200 modelo = './modelo_project_modelo_20200903_211740/modelo.h5' pesos_modelo = './modelo_project_modelo_20200903_211740/pesos.h5' cnn = load_model(modelo) cnn.load_weights(pesos_modelo) def open_img(): x = openfilename() img = Image.open(x) img = img.resize((300, 300), Image.ANTIALIAS) img = ImageTk.PhotoImage(img) panel = Label(root, image=img) panel.image = img panel.grid(row=2) img = keras.preprocessing.image.load_img( x, target_size=(altura, longitud) ) img_array = keras.preprocessing.image.img_to_array(img) img_array = tf.expand_dims(img_array, 0) predictions = cnn.predict(img_array) score = tf.nn.softmax(predictions[0]) class_names = ['CON MASCARILLA', 'SIN MASCARILLA'] messagebox.showinfo("CNN mask", "Imagen: {} \nTipo: {} \nPorcentaje de acierto: {:.2f}." .format(x, class_names[np.argmax(score)], 100 * np.max(score))) def openfilename(): filename = filedialog.askopenfilename(title='"Seleccione la imagen') return filename root = Tk() root.title("CNN Covid mask") root.geometry("350x350") root.resizable(width=True, height=True) btn = Button(root, text='Cargar una imagen', command=open_img).grid(row=1) root.mainloop()
{ "alphanum_fraction": 0.6856780735, "author": null, "avg_line_length": 27.6842105263, "converted": null, "ext": "py", "file": null, "hexsha": "6a22310329aa77b68564040c3994cc4b4f578736", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "52ea60cf534c75b0b1d78a54567b0fa5b90f797f", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "BryanManzano2016/CNN-mask---IA", "max_forks_repo_path": "interfaz_proyecto.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "52ea60cf534c75b0b1d78a54567b0fa5b90f797f", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "BryanManzano2016/CNN-mask---IA", "max_issues_repo_path": "interfaz_proyecto.py", "max_line_length": 93, "max_stars_count": null, "max_stars_repo_head_hexsha": "52ea60cf534c75b0b1d78a54567b0fa5b90f797f", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "BryanManzano2016/CNN-mask---IA", "max_stars_repo_path": "interfaz_proyecto.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 389, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1578 }
import vnmrjpy as vj import numpy as np from scipy.ndimage.filters import gaussian_filter, median_filter from vnmrjpy.core.utils import vprint import copy # for hiding zero-divide warnigns import warnings warnings.filterwarnings("ignore", category=RuntimeWarning) """ Generate fieldmap from a set of gradient echo images """ def make_fieldmap(varr, mask=None, selfmask=True, combine_receivers=True, \ method='triple_echo'): """Generate B0 map from gradient echo images. Args: varr -- input gradient echo image set with different echo time acquisitions at time dimension method -- triple_echo: see ref [1] mask -- ndarray of [0 or 1] with same shape of varr.data in spatial dims selfmask -- Boolean, set True to create mask based on magnitude data Return: fieldmap -- vj.varray with data attribute updated to the B0 map Refs: [1] Windischberger et al.: Robust Field Map Generation Using a Triple- Echo Acquisition, JMRI, (2004) """ if mask == None and selfmask==True: varr, mask = vj.func.mask(varr, mask_out=True) if method=='triple_echo': gaussian_kernel = _get_gaussian_kernel(varr) median_kernel = _get_median_kernel(varr) # checking for echos time_dim = varr.data.shape[3] # calcin milliseconds te = [float(i)*1000 for i in varr.pd['te']] phasedata = np.arctan2(np.imag(varr.data),np.real(varr.data)) magnitudedata = np.abs(varr.data) phasedata.astype('float32') phase_set = _make_phase_set(phasedata) d_set, c_set, residual_set = _calc_freq_shift(phase_set,te) indice_arr = _get_indice_map(residual_set) c_map = _get_from_set(c_set, indice_arr) res_map = _get_from_set(residual_set, indice_arr) # pre field map without filters and receivers not combined fieldmap = _get_from_set(d_set, indice_arr) # fieldmap processing fieldmap = _combine_receivers(fieldmap, res_map) fieldmap = _median_filter(fieldmap, kernel=median_kernel, mask=mask) fieldmap = _gaussian_filter(fieldmap, kernel=gaussian_kernel,mask=mask) # creating final varray varr.data = fieldmap varr.description = 'B0_map' return varr else: raise(Exception('Not implemented fieldmap generating method')) def _get_gaussian_kernel(varr): mult = 0.5 # means size in mm if type(varr.nifti_header) == type(None): varr = varr.set_nifti_header() affine = varr.nifti_header.get_qform() kernel = [1/i*mult for i in [affine[0,0],affine[1,1],affine[2,2]]] return kernel def _get_median_kernel(varr): slc_dim = varr.sdims.index('slice') kernel = [2,2,2] kernel[slc_dim] = 1 return tuple(kernel) def _make_phase_set(phasedata): """Return all possible phase wrapping combinations Args: phasedata -- numpy ndarray of dim (x,y,z, time, rcvrs) with phase data Return: phasedata_list -- list of possible phase data in all possible cases of phase wrapping Note: This phase ordering rule is also used in _get_fieldmap function """ # only implemented for 3 TE points p0 = phasedata[...,0,:] p1 = phasedata[...,1,:] p2 = phasedata[...,2,:] #See ref [1] in make_fieldmap() #case 1 case1 = phasedata phasedata_list = [case1] #case 2 case2 = np.stack([p0,p1+2*np.pi,p2+2*np.pi],axis=3) phasedata_list.append(case2) #case 3 case3 = np.stack([p0,p1-2*np.pi,p2-2*np.pi],axis=3) phasedata_list.append(case3) #case 4 case4 = np.stack([p0,p1,p2+2*np.pi],axis=3) phasedata_list.append(case4) #case 5 case5 = np.stack([p0,p1,p2-2*np.pi],axis=3) phasedata_list.append(case5) #case 6 case6 = np.stack([p0,p1+2*np.pi,p2],axis=3) phasedata_list.append(case6) #case 7 case7 = np.stack([p0,p1-2*np.pi,p2],axis=3) phasedata_list.append(case7) return phasedata_list def _calc_freq_shift(phase_set, te): """Calculate frequency shift at each point for each phase wrapping scenario Do linear regression of the form Phase(TE) = c + d * TE Args: phase_set -- list of phase sets in different phase wrapping cases te -- echo times in ms Return: d_set c_set residual_set """ d_set = [] c_set = [] residual_set = [] shape = phase_set[0].shape te = np.array(te,dtype=float) for num, phase in enumerate(phase_set): (x,y,z,t,rcvr) = phase.shape # reshape to accomodate polyfit vectorization phase = np.reshape(np.swapaxes(phase, 0,3),(t,-1)) out = np.polyfit(te, phase, 1, full=True) d,c = out[0] res = out[1] # reshape back to otiginal d = np.swapaxes(np.reshape(d,(1,y,z,x,rcvr)),0,3) c = np.swapaxes(np.reshape(c,(1,y,z,x,rcvr)),0,3) res = np.swapaxes(np.reshape(res,(1,y,z,x,rcvr)),0,3) # hack: just make the loss large where d is negative res[d < 0] = 10000 # append to list d_set.append(d) c_set.append(c) residual_set.append(res) return d_set, c_set, residual_set def _combine_receivers(arr, res_map): """Pick the value with the lowest residual from each chanel at a voxel""" min_ind = np.argmin(res_map, axis=4) arr = np.moveaxis(arr,[4,0,1,2,3],[0,1,2,3,4]) arr = np.choose(min_ind, arr) return np.expand_dims(arr,axis=4) def _get_indice_map(chisqr_set): """Find element with lowest chisqr at each voxel """ #make chisqr array of dims [x,y,z,0,rcvr,chisqr] chisqr_arr = np.stack(chisqr_set,axis=5) indice_arr = np.argmin(chisqr_arr,axis=5) return indice_arr def _get_from_set(par_set, indice_arr): """Extract values from set according to index array""" par_arr = np.stack(par_set,axis=0) par_map = np.choose(indice_arr, par_arr) return par_map def _median_filter(fieldmap, kernel=(3,3,3), mask=None): for rcvr in range(fieldmap.shape[4]): arr = copy.copy(fieldmap[...,0,rcvr]) fieldmap[...,0,rcvr] = median_filter(arr,size=kernel) return fieldmap def _gaussian_filter(fieldmap, kernel=(2,2,2), mask=None): for rcvr in range(fieldmap.shape[4]): arr = copy.copy(fieldmap[...,0,rcvr]) if type(mask) == type(None): fieldmap[...,0,rcvr] = gaussian_filter(arr,sigma=kernel) elif type(mask) != type(None): mask_rcvr = mask[...,0,rcvr] arr = gaussian_filter(arr * mask_rcvr,sigma=kernel) arr /= gaussian_filter(mask_rcvr,sigma=kernel) arr[mask_rcvr == 0] = 0 fieldmap[...,0,rcvr] = arr return fieldmap
{ "alphanum_fraction": 0.6383758429, "author": null, "avg_line_length": 32.9565217391, "converted": null, "ext": "py", "file": null, "hexsha": "25e1d2ce8c0312c5da1641ec6cbaa3b6df737651", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "48707a1000dc87e646e37c8bd686e695bd31a61e", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "hlatkydavid/vnmrjpy", "max_forks_repo_path": "vnmrjpy/func/fieldmap.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "48707a1000dc87e646e37c8bd686e695bd31a61e", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "hlatkydavid/vnmrjpy", "max_issues_repo_path": "vnmrjpy/func/fieldmap.py", "max_line_length": 80, "max_stars_count": null, "max_stars_repo_head_hexsha": "48707a1000dc87e646e37c8bd686e695bd31a61e", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "hlatkydavid/vnmrjpy", "max_stars_repo_path": "vnmrjpy/func/fieldmap.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1911, "path": null, "reason": "import numpy,from scipy", "repo": null, "save_path": null, "sha": null, "size": 6822 }
Quad Phone 2 on the Campus Payphones node Phone Number: (530) 7599256 Location: Sits not exactly in the Quad, but on the same block. Lives in the indented area next to the Memorial Union computer lab. Description: Its identical twin, Quad Phone 1, is to its right. Charge: 50 cents for local.
{ "alphanum_fraction": 0.7651006711, "author": null, "avg_line_length": 27.0909090909, "converted": null, "ext": "f", "file": null, "hexsha": "022a2499771e5345bf11425403e543ac771efdcd", "include": null, "lang": "FORTRAN", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "55088b2fe6a9d6c90590f090542e0c0e3c188c7d", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "voflo/Search", "max_forks_repo_path": "lab/davisWiki/Quad_Phone_2.f", "max_issues_count": null, "max_issues_repo_head_hexsha": "55088b2fe6a9d6c90590f090542e0c0e3c188c7d", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "voflo/Search", "max_issues_repo_path": "lab/davisWiki/Quad_Phone_2.f", "max_line_length": 130, "max_stars_count": null, "max_stars_repo_head_hexsha": "55088b2fe6a9d6c90590f090542e0c0e3c188c7d", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "voflo/Search", "max_stars_repo_path": "lab/davisWiki/Quad_Phone_2.f", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 77, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 298 }
# AUTOGENERATED! DO NOT EDIT! File to edit: nbs/02_transforms.fractal.ipynb (unless otherwise specified). __all__ = ['MandelBrotFractalTransform', 'JuliaFractalTransform'] # Cell import numpy as np from albumentations.core.transforms_interface import ImageOnlyTransform from PIL import Image # Cell class MandelBrotFractalTransform(ImageOnlyTransform): def __init__( self, n_iter: int = 10, deg: int = 2, delta: float = 0.01, always_apply: bool = False, p: float = 1.0, ): super(MandelBrotFractalTransform, self).__init__(always_apply, p) self.n_iter = n_iter self.deg = deg self.delta = delta self.F = lambda z, c, deg: np.power(z, deg) + c def apply(self, image, **params): image = Image.fromarray(image) if self.deg == 2: x_min, x_max = -2.0, 0.6 y_min, y_max = -1.2, 1.2 else: x_min, x_max = -2.5, 2 y_min, y_max = -1.5, 1.5 delta = self.delta re, im = np.mgrid[x_min:x_max:delta, y_min:y_max:delta] c = (re + 1j * im).T x0 = np.zeros_like(c) # x0 = 0 x = x0 btmleft = (-0.5, -0.5) topright = (0.5, 0.5) trap_size = (1.0, 1.0) fractal = np.zeros((*(np.abs(c).shape), 3)) trapped = np.zeros_like(np.abs(c)) with np.nditer(x, flags=["multi_index"], op_flags=["readwrite"]) as it: for point in it: for iter in range(self.n_iter): point[...] = self.F(point, c[it.multi_index], self.deg) # if this point was trapped previously then break if (abs(point) > 10) or trapped[it.multi_index] == 1: break # check if this point can be trapped if (point.real > btmleft[0] and point.real < topright[0]) and ( point.imag > btmleft[1] and point.imag < topright[1] ): trapped[it.multi_index] = 1 fractal[it.multi_index] = np.array( image.getpixel( ( int((point.real + 0.5) * image.width), int((point.imag + 0.5) * image.height), ) ) ) return fractal.astype(np.uint8) # Cell class JuliaFractalTransform(ImageOnlyTransform): def __init__( self, n_iter: int = 10, deg: int = 2, delta: float = 0.01, always_apply: bool = False, p: float = 1.0, ): super(JuliaFractalTransform, self).__init__(always_apply, p) self.n_iter = n_iter self.deg = deg self.delta = delta self.F = lambda z, c, deg: np.power(z, deg) + c def apply(self, image, **params): image = Image.fromarray(image) if self.deg == 2: x_min, x_max = -1.2, 1.2 y_min, y_max = -1.2, 1.2 else: x_min, x_max = -1.5, 1.5 y_min, y_max = -1.5, 1.5 delta = self.delta re, im = np.mgrid[x_min:x_max:delta, y_min:y_max:delta] grid = (re + 1j * im).T x = grid c = np.zeros_like(x) c.fill(0.3 + 0.6j) btmleft = (-0.5, -0.5) topright = (0.5, 0.5) trap_size = (1.0, 1.0) fractal = np.zeros((*(np.abs(c).shape), 3)) trapped = np.zeros_like(np.abs(c)) with np.nditer(x, flags=["multi_index"], op_flags=["readwrite"]) as it: for point in it: for iter in range(self.n_iter): point[...] = self.F(point, c[it.multi_index], self.deg) # if this point was trapped previously then break if (abs(point) > 10) or trapped[it.multi_index] == 1: break # check if this point can be trapped if (point.real > btmleft[0] and point.real < topright[0]) and ( point.imag > btmleft[1] and point.imag < topright[1] ): trapped[it.multi_index] = 1 pixel = ( int((point.real + 0.5) * (image.width - 1)), int((point.imag + 0.5) * (image.height - 1)), ) try: fractal[it.multi_index] = np.array(image.getpixel(pixel)) except: print( f"ERROR: pixel: ({int( ( point.real + 0.5 ) * image.width )}, {int( ( point.imag + 0.5 ) * image.height )}), point: ({point})" ) return fractal.astype(np.uint8)
{ "alphanum_fraction": 0.4702889703, "author": null, "avg_line_length": 33.4285714286, "converted": null, "ext": "py", "file": null, "hexsha": "57e461821f2fd61a69b152dd8b58b3fb354181a1", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "a43182f0c42896c5a9d5e519216ebc812cf7299d", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "satyajitghana/fractal_augmentation", "max_forks_repo_path": "fractal_augmentation/transforms/fractal.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "a43182f0c42896c5a9d5e519216ebc812cf7299d", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "satyajitghana/fractal_augmentation", "max_issues_repo_path": "fractal_augmentation/transforms/fractal.py", "max_line_length": 159, "max_stars_count": null, "max_stars_repo_head_hexsha": "a43182f0c42896c5a9d5e519216ebc812cf7299d", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "satyajitghana/fractal_augmentation", "max_stars_repo_path": "fractal_augmentation/transforms/fractal.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1290, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4914 }
\section{Decomposition 8: UC22, UC23 (application upload and statistics)} \subsection*{Selected architectural drivers} The functional drivers are: \begin{itemize} \item \emph{UC22}: Upload an application \item \emph{UC23}: Consult application statistics \end{itemize} % UC22: % 1. The primary actor indicates that he/she wants to upload an application. % IN CLIENT % % 2. The system asks the primary actor if he/she wants to update an existing application. % IN CLIENT % % 3. The primary actor provides his/her choice. % % IN CLIENT % % 4. If the primary actor indicated that he/she wants to update an existing application, % the system composes an overview of applications uploaded by the primary actor % (e.g., as a list or table), presents this, and requests the primary actor to indicate which application to update. % % ApplicationProviderClient -> ApplicationProviderFacade: interface Applications: List<Application> getApplicationsOfApplicationProvider(int applicationProviderID) % ApplicationProviderFacade -> ApplicationManager: interface FrontEndAppRequests: List<Application> getApplicationsOfApplicationProvider(int applicationProviderID) % ApplicationManager -> OtherDataDB: interface DBAppMgmt: List<Application> getApplicationsOfApplicationProvider(int applicationProviderID) % % 5. The primary actor chooses the application to be updated. % IN CLIENT % % 6. The system asks the primary actor if this is an update that should be automatically activated for existing subscriptions. % IN CLIENT % % 7. The primary actor provides his/her choice. % IN CLIENT % % 8. If the primary actor has indicated that existing subscriptions should be automatically updated, the system requests the versions (or range of versions) that should be automatically updated. % IN CLIENT % % 9. The primary actor provides the requested information. % IN CLIENT % % 10. The system asks the primary actor to upload the application code, a description for display to customer organisations, and meta-data (such as the version number and subscription price). % IN CLIENT % % 11. The primary actor uploads the requested information. % ApplicationProviderClient -> ApplicationProviderFacade: interface Applications: void uploadApplication(int applicationProviderID, ApplicationCode code, string description, Map<string, string> metaData, List<Version> versionsToUpdate) % ApplicationProviderFacade -> ApplicationManagementLogic: interface FrontEndAppRequests: void uploadApplication(int applicationProviderID, ApplicationCode code, string description, Map<string, string> metaData, List<Version> versionsToUpdate) % ApplicationManagementLogic -> OtherDataDB: interface DBAppMgmt: int createNewApplication(int applicationProviderID, ApplicationCode code, string description, Map<string, string> metaData, List<Version> versionsToUpdate) % % 12. The system performs automated application checks. % % ApplicationManagementLogic -> ApplicationContainerManager: interface AppInstanceMgmt: void testApplication(int applicationID, ApplicationCode code) % % 13. If the checks were successful, the system sends a notification to the primary actor (Include: UC15 : Send notification). % % ApplicationContainerManager -> ApplicationManagementLogic: interface ApplicationTesting: void applicationTestsSuccessful(int applicationID) % ApplicationContainerManager-> ApplicationManagementLogic: interface ApplicationTesting: void applicationTestsUnsuccessful(int applicationID) % % ApplicationManageMentLogic -> NotificationHandler: interface Notify: notify(int userID) app provider % ApplicationManageMentLogic -> NotificationHandler: interface Notify: notify(int userID) sys admin % % 14. The system makes the application available in the application store. In case of an updated application, it replaces the previous versions of the application. % % ApplicationManagementLogic -> OtherDataDB: interface DBAppMgmt: void updateApplicationAvailable(int applicationID) % % 15. In case of an updated application that should automatically be activated for existing subscriptions (cf. step 6), % the system updates the existing subscriptions of the application to the new version (Include: UC17 : Activate an application). % % ApplicationManagementLogic -> OtherDataDB: interface DBAppMgmt: List<int> getApplicationInstancesToUpdate(int applicationID) % ApplicationManagementLogic -> OtherDataDB: interface DBAppMgmt: ApplicationCode getApplicationCode(int applicationID) % ApplicationManagementLogic -> ApplicationContainerManager: interface AppInstanceMgmt: void updateApplicationInstances(List<int> applicationInstanceIDs, ApplicationCode code) % ApplicationManagementLogic -> OtherDataDB: interface DBAppMgmt: void updateApplicationInstancesOfApplication(int applicationID) % % ApplicationManagementLogic -> ApplicationContainerManager: interface AppInstanceMgmt: void createApplicationInstance(ApplicationInstance instance, ApplicationCode code) % UC23: % 1. The primary actor indicates that he/she wants to an overview of uploaded applications. % % ApplicationProviderClient -> ApplicationProviderFacade: interface Applications: Map<Applications, Map<string, string>> getApplicationsWithStatsForApplicationProvider(int applicationProviderID) % % 2. The system retrieves all applications uploaded by the primary actor, and information about the amount of subscribers for that application. It presents this overview to the primary actor. % % ApplicationProviderFacade -> ApplicationManager: interface FrontEndAppRequests: Map<Applications, Map<string, string>> getApplicationsWithStatsForApplicationProvider(int applicationProviderID) % ApplicationManager -> OtherDataDB: interface DBAppMgmt: Map<Applications, Map<string, string>> getApplicationsWithStatsForApplicationProvider(int applicationProviderID) % % 3. The primary actor selects an application. % % ApplicationProviderClient -> ApplicationProviderFacade: interface Applications: Map<string, string> getApplicationDetailedStats(int applicationProviderID, int applicationID) % % 4. If the selected application is an approved application, the system presents detailed statistics including the amount of subscribers. % % ApplicationProviderFacade -> ApplicationManager: interface FrontEndAppRequests: Map<string, string> getApplicationDetailedStats(int applicationID) % ApplicationManager -> OtherDataDB: interface DBAppMgmt: Map<string, string> getApplicationDetailedStats(int applicationID) \subsection*{Interfaces for child modules} This section lists new interfaces assigned to the components defined in the section above. Detailed information about each interface and its methods can be found under chapter \ref{ch:elements-datatypes}. \subsubsection{ApplicationProviderFacade} \begin{itemize} \item Applications \end{itemize} \subsubsection{ApplicationManagementLogic} \begin{itemize} \item ApplicationTesting \end{itemize} \subsection*{New data types} This section lists the new data types introduced during this decomposition. \begin{itemize} \item Version \item ApplicationCode \end{itemize}
{ "alphanum_fraction": 0.739755663, "author": null, "avg_line_length": 64.9421487603, "converted": null, "ext": "tex", "file": null, "hexsha": "7991d9dcbbfc822395574d981f0492f1904eded3", "include": null, "lang": "TeX", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "56f14be5801eb7c4a8dcb53f552ca36ffbc13066", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "arminnh/software-architecture-assignment", "max_forks_repo_path": "part-2b/report/decomposition-8.tex", "max_issues_count": null, "max_issues_repo_head_hexsha": "56f14be5801eb7c4a8dcb53f552ca36ffbc13066", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "arminnh/software-architecture-assignment", "max_issues_repo_path": "part-2b/report/decomposition-8.tex", "max_line_length": 255, "max_stars_count": 1, "max_stars_repo_head_hexsha": "56f14be5801eb7c4a8dcb53f552ca36ffbc13066", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "arminnh/software-architecture-assignment", "max_stars_repo_path": "part-2b/report/decomposition-8.tex", "max_stars_repo_stars_event_max_datetime": "2017-12-16T08:12:00.000Z", "max_stars_repo_stars_event_min_datetime": "2017-12-16T08:12:00.000Z", "num_tokens": 1567, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 7858 }
# ----------------------------------------------------------------------------- # # -*- coding: utf-8 -*- # # phlox-libdc1394/dc1394/frame.py # # Copyright (C) 2016, by Matthias Yang Chen <matthias_cy@outlook.com> # All rights reserved. # # phlox-libdc1394 is free software: you can redistribute it and/or modify it # under the terms of the GNU Lesser General Public License as # published by the Free Software Foundation, either version 3 of the # License, or (at your option) any later version. # # phlox-libdc1394 is distributed in the hope that it will be useful, but # WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU # Lesser General Public License for more details. # # You should have received a copy of the GNU Lesser General Public # License along with phlox-libdc1394. If not, # see <http://www.gnu.org/licenses/>. # ----------------------------------------------------------------------------- from __future__ import division, print_function from __future__ import absolute_import, unicode_literals from ctypes import ARRAY, c_byte from numpy import ndarray from .core import * __all__ = ['Frame'] class Frame(ndarray): """ A frame returned by the camera. All metadata are retained as attributes of the resulting image. """ _cam = None _frame = None def __new__(cls, camera, frame): """ Convert a dc1394 frame into a Frame instance. :param camera: :param frame: :return: """ dtype = ARRAY(c_byte, frame.contents.image_bytes) buf = dtype.from_address(frame.contents.image) width, height = frame.contents.size pixels = width * height endian = frame.contents.little_endian and '<' or '>' type_str = '%su%i' % (endian, frame.contents.image_bytes / pixels) img = ndarray.__new__(cls, shape=(height, width), dtype=type_str, buffer=buf) img.frame_id = frame.contents.id img.frames_behind = frame.contents.frames_behind img.position = frame.contents.position img.packet_size = frame.contents.packet_size img.packets_per_frame = frame.contents.packet_per_frame img.timestamp = frame.contents.timestamp img.video_mode = video_modes[frame.contents.video_mode] img.data_depth = frame.contents.data_depth img.color_coding = color_codings[frame.contents.color_coding] img.color_filter = frame.contents.color_filter img.yuv_byte_order = frame.contents.yuv_byte_order img.stride = frame.contents.stride # save camera and frame for enqueue() img._frame = frame img._cam = camera return img def __array_finalize__(self, img): """ Finalize the new Image class array. If called with an image object, inherit the properties of that image. """ if img is None: return # do not inherit _frame and _cam since we also get called on copy() # and should not hold references to the frame in this case for key in ["position", "color_coding", "color_filter", "yuv_byte_order", "stride", "packet_size", "packets_per_frame", "timestamp", "frames_behind", "frame_id", "data_depth", "video_mode"]: setattr(self, key, getattr(img, key, None)) def enqueue(self): """ Returns a frame to the ring buffer once it has been used. This method is also called implicitly on ``del``. Only call this method on the original frame obtained from Camera.dequeue` and not on its views, new-from-templates or copies. Otherwise an AttributeError will be raised. """ if not hasattr(self, "_frame"): # or self.base is not None: raise AttributeError("can only enqueue the original frame") if self._frame is not None: dll.dc1394_capture_enqueue(self._cam, self._frame) self._frame = None self._cam = None # from contextlib iport closing # with closing(camera.dequeue()) as im: # do stuff with im close = enqueue def __del__(self): try: self.enqueue() except AttributeError: pass @property def corrupt(self): """ Whether this frame corrupt. Returns ``True`` if the given frame has been detected to be corrupt (missing data, corrupted data, overrun buffer, etc.) and ``False`` otherwise. .. note:: Certain types of corruption may go undetected in which case ``False`` will be returned erroneously. The ability to detect corruption also varies between platforms. .. note:: Corrupt frames still need to be enqueued with `enqueue` when no longer needed by the user. """ return bool(dll.dc1394_capture_is_frame_corrupt(self._cam, self._frame)) def to_rgb(self): """ Convert the image to an RGB image. Array shape is: (image.shape[0], image.shape[1], 3) Uses the dc1394_convert_to_RGB() function for the conversion. """ res = ndarray(3 * self.size, dtype='u1') shape = self.shape inp = ndarray(shape=len(self.data), buffer=self.data, dtype='u1') dll.dc1394_convert_to_RGB8(inp, res, shape[1], shape[0], self.yuv_byte_order, self.color_coding, self.data_depth) res.shape = shape[0], shape[1], 3 return res def to_mono8(self): """ Convert he image to 8 bit gray scale. Uses the dc1394_convert_to_MONO8() function """ res = ndarray(self.size, dtype='u1') shape = self.shape inp = ndarray(shape=len(self.data), buffer=self.data, dtype='u1') dll.dc1394_convert_to_MONO8(inp, res, shape[1], shape[0], self.yuv_byte_order, self.color_coding, self.data_depth) res.shape = shape return res def to_yuv422(self): """ Convert he image to YUV422 color format. Uses the dc1394_convert_to_YUV422() function """ res = ndarray(self.size, dtype='u1') shape = self.shape inp = ndarray(shape=len(self.data), buffer=self.data, dtype='u1') dll.dc1394_convert_to_YUV422(inp, res, shape[1], shape[0], self.yuv_byte_order, self.color_coding, self.data_depth) return ndarray(shape=shape, buffer=res.data, dtype='u2')
{ "alphanum_fraction": 0.6030410393, "author": null, "avg_line_length": 38.0561797753, "converted": null, "ext": "py", "file": null, "hexsha": "22df73cb320628fed7eb171e06fab821cbb9e8f4", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "f312569ec983b5f27c75846b34debc04fe7bdf98", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "jardinier/phlox", "max_forks_repo_path": "PhloxAR/dc1394/frame.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "f312569ec983b5f27c75846b34debc04fe7bdf98", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "jardinier/phlox", "max_issues_repo_path": "PhloxAR/dc1394/frame.py", "max_line_length": 85, "max_stars_count": 1, "max_stars_repo_head_hexsha": "f312569ec983b5f27c75846b34debc04fe7bdf98", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "jardinier/phlox", "max_stars_repo_path": "PhloxAR/dc1394/frame.py", "max_stars_repo_stars_event_max_datetime": "2016-05-22T00:12:14.000Z", "max_stars_repo_stars_event_min_datetime": "2016-05-22T00:12:14.000Z", "num_tokens": 1478, "path": null, "reason": "from numpy", "repo": null, "save_path": null, "sha": null, "size": 6774 }
#' @title save_data #' @description This function takes either a dataframe or all of the data you've filtered, and rolls it up into a csv and/ #' or a shapefile for continued analysis #' @param df This is a dataframe that you want to save in some other format. If a spatial format is selected #' (e.g. shapefile), it must have LATITUDE and LONGITUDE fields #' @param filename default is \code{NULL}. This will be the prefix of your filename #' @param df.crs This is the CRS value for your dataframe. This should be the reference system that your data is known to be in. #' The default value \code{"EPSG:4326"} is WGS84 and is appropriate for most data collected using a GPS. #' @param db default is \code{NULL}. This identifies the dataset you are working #' with. #' @param req.coords default is TRUE. This filters out records without values for LATITUDE or #' LONGITUDE. The function aborts if req.coords=TRUE and no records remain. #' @param lat.field the default is \code{"LATITUDE"}. the name of the field holding latitude values (in decimal degrees) #' @param lon.field the default is \code{"LONGITUDE"}. the name of the field holding longitude values (in decimal degrees) #' @param formats This is a vector of the formats in which you would like to save the current data, #' including "raw" for a (local) dataframe, "sp" for a (local) SpatialPointsDataFrame, #' "shp" for a shapefile or "csv" (both written to the wd). The raw and sp objects will #' just have the name specified by filename, while the csv and shapefiles, since they're #' written externally also get a timestamp. #' @param env This the the environment you want this function to work in. The #' default value is \code{.GlobalEnv}. #' @family general_use #' @author Mike McMahon, \email{Mike.McMahon@@dfo-mpo.gc.ca} #' @export save_data <- function(db = NULL, df= NULL, filename = NULL, df.crs = "EPSG:4326", req.coords=TRUE, lat.field = "LATITUDE", lon.field = "LONGITUDE", formats = c('csv', 'shp'), env=.GlobalEnv){ if (req.coords == FALSE & 'shp' %in% formats) warning("\n","Since req.coords = FALSE, not all of the records necessarily have positions and will not be visible in your shapefile") if (is.null(df)) { df = summarize_catches(db=ds_all[[.GlobalEnv$db]]$db, valid.coords = req.coords, env=env, drop.na.cols = F) if (is.null(df)){ cat("\n","No records to save") return(NULL) } } if (is.null(filename)) { name = match.call()[2] }else { name = filename } name = gsub('()','',name) name = gsub('\\.','',name) ts = format(Sys.time(), "%Y%m%d_%H%M") fn = paste(name,"_",ts,sep="" ) #id posix and date fields df=data.frame(lapply(df, function(x) if(inherits(x, "POSIXct")|inherits(x, "Date")) as.Date(strftime(x, format="%Y-%m-%d")) else x)) if ('raw' %in% formats) assign(paste0("raw_",name), df, envir = env) if ('shp' %in% formats | 'sp' %in% formats){ df.sp = Mar.utils::df_qc_spatial(df, lat.field, lon.field) df.sp = sp::SpatialPointsDataFrame( coords = df.sp[, c(lon.field, lat.field)], data = df.sp, proj4string = sp::CRS(SRS_string=df.crs) ) if (nrow(df.sp@data) != nrow(df)) { cat("\n",paste0(nrow(df)-nrow(df.sp@data), " records were lost due to invalid coordinates")) } if ('sp' %in% formats) assign(paste0("sp_",name), df.sp, envir = env) if ('shp' %in% formats){ df.sp = Mar.utils::prepare_shape_fields(df.sp) rgdal::writeOGR(df.sp, ".", fn, driver="ESRI Shapefile", overwrite_layer=TRUE) cat(paste0("\nWrote file to: ",file.path(getwd(),fn))) } } if ('csv' %in% formats){ utils::write.csv(df, paste0(fn,".csv")) } if ('rds' %in% formats){ saveRDS(df, paste0(fn,".rds")) } return(invisible(df)) }
{ "alphanum_fraction": 0.6494579246, "author": null, "avg_line_length": 49.6666666667, "converted": null, "ext": "r", "file": null, "hexsha": "0e2f97f4b488fcb50e53247cc74ed2f7c0f24718", "include": null, "lang": "R", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "19f31cdfc11033dacad547fae958d553d93cec7d", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "Maritimes/Mar.datawrangling", "max_forks_repo_path": "R/save_data.r", "max_issues_count": 1, "max_issues_repo_head_hexsha": "19f31cdfc11033dacad547fae958d553d93cec7d", "max_issues_repo_issues_event_max_datetime": "2017-11-08T19:39:31.000Z", "max_issues_repo_issues_event_min_datetime": "2017-11-08T19:39:31.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "Maritimes/Mar.datawrangling", "max_issues_repo_path": "R/save_data.r", "max_line_length": 134, "max_stars_count": 2, "max_stars_repo_head_hexsha": "19f31cdfc11033dacad547fae958d553d93cec7d", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "Maritimes/Mar.datawrangling", "max_stars_repo_path": "R/save_data.r", "max_stars_repo_stars_event_max_datetime": "2020-03-11T21:01:57.000Z", "max_stars_repo_stars_event_min_datetime": "2019-02-07T23:42:10.000Z", "num_tokens": 1080, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 3874 }
import sys import numpy as np import torch from tqdm import trange from atom_joggling.utils import ROOT, mean, parser # Method specific options # Mixup parameter of the Beta distribution parser.add_argument("--alpha", default=0.75, type=float) # weighting factor for the unlabeled loss term parser.add_argument("--lambda-u", default=75, type=float) # temperature used during sharpening of the pseudo-label distribution parser.add_argument("--T", default=0.5, type=float) # mixmatch used a default of 1024 but CIFAR10 is a way bigger dataset so 128 seems good parser.add_argument( "--train-iterations", type=int, default=128, help="Number of batches per epoch" ) args, _ = parser.parse_known_args() # add time stamp if no custom out_dir was specified if not args.resume and not args.out_dir: out_dir = f"{ROOT}/runs/mixup/iters={args.train_iterations}_T={args.T}" out_dir += f"alpha={args.alpha}-lambdaU={args.lambda_u}" out_dir += "_robust" if args.robust else "" args.out_dir = out_dir def train_with_mixup( labeled_loader, unlabeled_loader, model, optimizer, criterion, verbose=True ) -> tuple: """Train a model with mixup by randomly sampling linear combinations of unlabeled and unlabeled inputs as well as targets. Uses model-generated pseudo-labels to compute interpolated targets. Args: labeled_loader (torch.utils.data.DataLoader): labeled data unlabeled_loader (torch.utils.data.DataLoader): unlabeled data model (nn.Module): the instantiated model optimizer (torch.optim.Optimizer): optimizer criterion: loss function that computes labeled, unlabeled and combined losses Returns: [loss, Lx, Lu]: 3-tuple of total, labeled and unlabeled losses """ losses = {"total": [], "loss_u": [], "loss_x": []} labeled_train_iter = iter(labeled_loader) unlabeled_train_iter = iter(unlabeled_loader) model.train() # file=sys.stdout (default stderr) prevents print order issues by using # same stream as print https://stackoverflow.com/a/45265707 for _ in trange( args.train_iterations, desc="Batches:", file=sys.stdout, disable=not verbose ): try: inputs_x, targets_x, *_ = next(labeled_train_iter) except StopIteration: labeled_train_iter = iter(labeled_loader) inputs_x, targets_x, *_ = next(labeled_train_iter) try: inputs_u, *_ = next(unlabeled_train_iter) except StopIteration: unlabeled_train_iter = iter(unlabeled_loader) inputs_u, *_ = next(unlabeled_train_iter) batch_size = targets_x.size(0) # Transform labels to one-hot targets_x = torch.nn.functional.one_hot(targets_x) crys_fea_x = model.material_nn(*inputs_x) crys_fea_u = model.material_nn(*inputs_u) with torch.no_grad(): # compute guessed labels of unlabeled samples outputs_u = model(*inputs_u) # sharpen the model's output distribution (for entropy minimization) proba_u = outputs_u.softmax(dim=1) pt = proba_u ** (1 / args.T) targets_u = pt / pt.sum(dim=1, keepdim=True) targets_u = targets_u.detach() # mixup all_inputs = torch.cat([crys_fea_x, crys_fea_u], dim=0) all_targets = torch.cat([targets_x, targets_u], dim=0) ell = np.random.beta(args.alpha, args.alpha) ell = max(ell, 1 - ell) idx = torch.randperm(all_inputs.size(0)) input_a, input_b = all_inputs, all_inputs[idx] target_a, target_b = all_targets, all_targets[idx] mixed_input = ell * input_a + (1 - ell) * input_b mixed_target = ell * target_a + (1 - ell) * target_b # interleave labeled and unlabeled samples between batches to # get correct batchnorm calculation mixed_input = list(torch.split(mixed_input, batch_size)) mixed_input = interleave(mixed_input, batch_size) logits = [model.output_nn(mixed_input[0])] for input in mixed_input[1:]: logits.append(model.output_nn(input)) # put interleaved samples back logits = interleave(logits, batch_size) logits_x = logits[0] logits_u = torch.cat(logits[1:], dim=0) Lx, Lu, u_ramp = criterion( logits_x, mixed_target[:batch_size], logits_u, mixed_target[batch_size:] ) loss = Lx + u_ramp * Lu # record loss losses["total"].append(loss.item()) losses["loss_x"].append(Lx.item()) losses["loss_u"].append(Lu.item()) # update model weights optimizer.zero_grad() loss.backward() optimizer.step() # dicts are insertion ordered as of Python 3.6 losses = (mean(x) for x in losses.values()) return (*losses, u_ramp) @torch.no_grad() def validate_mixup(val_loader, model, criterion) -> tuple: losses, avg_acc = [], [] # switch to evaluate mode model.eval() for inputs, targets, *_ in val_loader: # compute output outputs = model(*inputs) loss = criterion(outputs, targets) # measure accuracy and record loss acc = (outputs.argmax(1) == targets).float().mean() losses.append(loss.item()) avg_acc.append(acc.item()) return mean(losses), mean(avg_acc) def linear_rampup(current: int, rampup_length: int) -> float: """Linearly ramps up the unlabeled loss with epoch count. As the pseudo-labels become more accurate, they can be be weighted more. Args: current (float): current loss weighting factor rampup_length (ing, optional): number of epochs until rampup completes. Defaults to args.epochs. Returns: float: increasing weighting factor for the unlabeled loss """ if rampup_length == 0: return 1.0 else: current /= rampup_length current = max(0.0, min(current, 1.0)) return current class SemiLoss: def __init__(self, u_ramp_length: int, ramp_start: int = 0) -> None: self.u_ramp_length = u_ramp_length self.ramp_count = ramp_start def __call__(self, preds_x, targets_x, preds_u, targets_u) -> tuple: self.ramp_count += 1 probs_u = preds_u.softmax(dim=1) Lx = -(preds_x.log_softmax(dim=1) * targets_x).sum(dim=1).mean() Lu = (probs_u - targets_u).pow(2).mean() # MSE ramp = linear_rampup(self.ramp_count, self.u_ramp_length) return Lx, Lu, args.lambda_u * ramp def interleave_offsets(batch_size: int, nu: int) -> list: groups = [batch_size // (nu + 1)] * (nu + 1) for x in range(batch_size - sum(groups)): groups[-x - 1] += 1 offsets = [0] for g in groups: offsets.append(offsets[-1] + g) assert offsets[-1] == batch_size return offsets def interleave(xy: list, batch_size: int) -> list: """Interleave labeled and unlabeled samples between batches to get correct batch normalization calculation. Args: xy (tuple): list of inputs or targets split by batch size batch_size (int): batch size Returns: [list]: list of interleaved inputs or targets """ nu = len(xy) - 1 offsets = interleave_offsets(batch_size, nu) xy = [[v[offsets[p] : offsets[p + 1]] for p in range(nu + 1)] for v in xy] for i in range(1, nu + 1): xy[0][i], xy[i][i] = xy[i][i], xy[0][i] return [torch.cat(v, dim=0) for v in xy]
{ "alphanum_fraction": 0.6470119522, "author": null, "avg_line_length": 33.1718061674, "converted": null, "ext": "py", "file": null, "hexsha": "37d4ebb25acea6ac26d7e6196c6ced81b9b07121", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-10-05T11:47:58.000Z", "max_forks_repo_forks_event_min_datetime": "2021-10-05T11:47:58.000Z", "max_forks_repo_head_hexsha": "5d9b19f136634b5d4d7a92a5b815510902b7da54", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "janosh/atom-joggling", "max_forks_repo_path": "atom_joggling/mixup.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "5d9b19f136634b5d4d7a92a5b815510902b7da54", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "janosh/atom-joggling", "max_issues_repo_path": "atom_joggling/mixup.py", "max_line_length": 87, "max_stars_count": 1, "max_stars_repo_head_hexsha": "5d9b19f136634b5d4d7a92a5b815510902b7da54", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "janosh/atom-joggling", "max_stars_repo_path": "atom_joggling/mixup.py", "max_stars_repo_stars_event_max_datetime": "2021-10-05T11:47:56.000Z", "max_stars_repo_stars_event_min_datetime": "2021-10-05T11:47:56.000Z", "num_tokens": 1879, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 7530 }
[STATEMENT] lemma order_dense_order_ow_transfer[transfer_rule]: assumes [transfer_rule]: "bi_unique A" "right_total A" "bi_unique B" "right_total B" shows "( rel_set A ===> (A ===> A ===> (=)) ===> (A ===> A ===> (=)) ===> rel_set B ===> (B ===> B ===> (=)) ===> (B ===> B ===> (=)) ===> (=) ) order_dense_order_ow order_dense_order_ow" [PROOF STATE] proof (prove) goal (1 subgoal): 1. (rel_set A ===> (A ===> A ===> (=)) ===> (A ===> A ===> (=)) ===> rel_set B ===> (B ===> B ===> (=)) ===> (B ===> B ===> (=)) ===> (=)) order_dense_order_ow order_dense_order_ow [PROOF STEP] by (ow_locale_transfer locale_defs: order_dense_order_ow_def)
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": "Types_To_Sets_Extension_Examples_TTS_Foundations_Orders_Set_Simple_Orders", "hexsha": null, "include": null, "lang": null, "length": 1, "llama_tokens": 284, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
import scanpy as sc import muon as mu import numpy as np ## VIASH START par = { 'input': 'resources_test/pbmc_1k_protein_v3/pbmc_1k_protein_v3_filtered_feature_bc_matrix.h5mu', 'modality': ['rna'], 'output': 'output.h5mu', 'var_name_filter': 'filter_with_hvg', 'do_subset': False, 'flavor': 'seurat', 'n_top_genes': 123, 'min_mean': 0.0125, 'max_mean': 3.0, 'min_disp': 0.5, 'span': 0.3, 'n_bins': 20, 'varm_name': 'hvg' } ## VIASH END mdata = mu.read_h5mu(par["input"]) mdata.var_names_make_unique() for mod in par['modality']: print(f"Processing modality '{mod}'") data = mdata.mod[mod] #sc.pp.log1p(data) print(f" Unfiltered data: {data}") print(" Computing hvg") # construct arguments hvg_args = { 'adata': data, 'n_top_genes': par["n_top_genes"], 'min_mean': par["min_mean"], 'max_mean': par["max_mean"], 'min_disp': par["min_disp"], 'span': par["span"], 'n_bins': par["n_bins"], 'flavor': par["flavor"], 'subset': False, 'inplace': False } # only add parameter if it's passed if par.get("max_disp", None) is not None: hvg_args["max_disp"] = par["max_disp"] if par.get("obs_batch_key", None) is not None: hvg_args["batch_key"] = par["obs_batch_key"] # call function try: out = sc.pp.highly_variable_genes(**hvg_args) out.index = data.var.index except ValueError as err: if str(err) == "cannot specify integer `bins` when input data contains infinity": err.args = ("Cannot specify integer `bins` when input data contains infinity. Perhaps input data has not been log normalized?",) raise err print(" Storing output into .var") if par.get("var_name_filter", None) is not None: data.var[par["var_name_filter"]] = out["highly_variable"] if par.get("varm_name", None) is not None: # drop mean_bin as muon/anndata doesn't support tuples data.varm[par["varm_name"]] = out.drop("mean_bin", axis=1) if par["do_subset"]: keep_feats = np.ravel(data.var[par["var_name_filter"]]) mdata.mod[mod] = data[:,keep_feats] # # can we assume execution_log exists? # if mdata.uns is None or "execution_log" not in mdata.uns: # mdata.uns["execution_log"] = [] # # store new entry # new_entry = {"component": meta["functionality_name"], "params": par} # mdata.uns["execution_log"].append(new_entry) print("Writing h5mu to file") mdata.write_h5mu(par["output"])
{ "alphanum_fraction": 0.6294695481, "author": null, "avg_line_length": 29.5930232558, "converted": null, "ext": "py", "file": null, "hexsha": "cfed615c93813dc24fbc5ba55b533312a1aae9d2", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "1e030961303b21b4351503cdc5f0a25d2879d094", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "openpipeline-bio/openpipeline", "max_forks_repo_path": "src/filter/filter_with_hvg/script.py", "max_issues_count": 4, "max_issues_repo_head_hexsha": "1e030961303b21b4351503cdc5f0a25d2879d094", "max_issues_repo_issues_event_max_datetime": "2022-02-02T11:05:20.000Z", "max_issues_repo_issues_event_min_datetime": "2021-10-20T12:58:52.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "openpipeline-bio/openpipeline", "max_issues_repo_path": "src/filter/filter_with_hvg/script.py", "max_line_length": 140, "max_stars_count": 2, "max_stars_repo_head_hexsha": "1e030961303b21b4351503cdc5f0a25d2879d094", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "openpipeline-bio/openpipeline", "max_stars_repo_path": "src/filter/filter_with_hvg/script.py", "max_stars_repo_stars_event_max_datetime": "2022-02-11T12:46:43.000Z", "max_stars_repo_stars_event_min_datetime": "2021-11-10T13:16:09.000Z", "num_tokens": 740, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2545 }
#include <stdlib.h> /* srand, rand */ #include <iostream> #include <fstream> // creating (.csv) files #include <vector> #include <cstring> #include <string> #include <functional> #include <armadillo> // http://arma.sourceforge.net/docs.html #include <thread> /* std::this_thread::sleep_for */ #include "EvolutionaryAlgorithm.hpp" // NOT USED?? // enum events{ // BASE_RATE = 0, // MUTA_RATE, // PRED_RATE, // PART_RATE, // GENO_RATE, // EV_A_RATE, // }; //arma::rowvec semiFinalsAndFinals[SEMI_FINALS][TOURNMENT_RATE]; //arma::rowvec indvFinals[TOURNMENT_RATE]; void EvolutionaryAlgorithm::saveBestIndvParamsCSV(){ std::fstream myFile; myFile.open("constValues.cfg", std::ios_base::app); myFile << parametersList[0] << " = " << population(bestFitIndex,0); for (size_t i = 1; i < NB_PARAMETERS; i++) myFile << std::endl << parametersList[i] << " = " << population(bestFitIndex,i); //printf("k%sk\n", parametersList[i]); myFile.close(); } // 1st step: initialize population void EvolutionaryAlgorithm::initializePop(int tournmentType) { double minVal,maxVal; std::cout << "INITIALIZING POPULATION\n"; // for (int i = 0; i < POPULATION_SIZE; i++) { // for (int j = 0; j < NB_PARAMETERS; j++) { // population[i][j] = (double) (rand() % MAX_PARAM_VALUE); // number range = [0, MAX_PARAM_VALUE[ // } // } population.randu(); // initialize with values between 0 and 1 if(tournmentType == INITIALS) population *= MAX_PARAM_VALUE; // if a normal EA is taking place, just multiply the population by a givern value else{ int middle = TOURNMENT_RATE + (POPULATION_SIZE - TOURNMENT_RATE)/2; fillInitialsWithBests(tournmentType); for(int j = 0;j< NB_PARAMETERS;j++){ // Min and max values are defined for each col, and converting the current col values to a number between this constraints. // this starts after the first TOURNMENT_RATE individuls, and the remaining is divided into two pars, one with biased rand values, and other without minVal = MIN_MULT * bestTournmentIndv[tournmentType].col(j).min(); maxVal = MAX_MULT * bestTournmentIndv[tournmentType].col(j).max(); for(int i = TOURNMENT_RATE;i<middle;i++){ population(i,j) = minVal + (maxVal - minVal)*(population(i,j)); } } for(int i = middle;i < POPULATION_SIZE;i++) population.row(i) *= MAX_PARAM_VALUE; } population.print("Population matrix initialized:"); return; } // Function to normalize fitness (values between 0 and 1) void EvolutionaryAlgorithm::normalizeFitness(double * normalizedFitness) { if (normalizedFitness == NULL) return; for (int i = 0; i < POPULATION_SIZE; i++) normalizedFitness[i] = (fitness[i]-bestFitness)/(worstFitness-bestFitness); return; } // 2nd step: evaluate population (calculate fitness) void EvolutionaryAlgorithm::evaluatePop() { std::cout << "EVALUATING POPULATION\n"; nbNoImprovementGens++; // we begin considering there was no improvement in the generation pthread_mutex_lock(&mutex); remainingFitnessToCalc = POPULATION_SIZE; pthread_mutex_unlock(&mutex); for (int i = 0; i<POPULATION_SIZE; i++){ m_eaGameControler.startGame(i,arma::conv_to<std::vector<double>>::from(population.row(i))); } for (int i = 0; i< POPULATION_SIZE; i++){ pthread_mutex_lock(&mutex2); } for (int i = 0; i < POPULATION_SIZE; i++) { // example of fitness calculation -> CHANGE!!!! // fitness[i] = abs(terminoReal - terminoEsperado)/terminoEsperado if (fitness[i] < bestFitness){ // searching for the max fitnnes from new generation bestFitness = fitness[i]; bestFitIndex = i; nbNoImprovementGens = 0; // this generation shows improvement -> reset counter } else if (fitness[i] > worstFitness){ worstFitness = fitness[i]; worstFitIndex = i; } // printf("fitness[%d] %lf\n", i, fitness[i]); } printf("BEST FITNESS: %lf - INDEX: %d\n", bestFitness, bestFitIndex); printf("WORST FITNESS: %lf - INDEX: %d\n", worstFitness, worstFitIndex); return; } void EvolutionaryAlgorithm::crossover(int indv, arma::rowvec parent1, arma::rowvec parent2) { // for(int j = 0; j < NB_PARAMETERS; j++){ // population[i][j] = (parent1[j] + parent2[j])/2.0; // } population.row(indv) = (parent1 + parent2)/2.0; return; } void EvolutionaryAlgorithm::elitism() { std::cout << "ELITISM\n"; arma::rowvec bestIndv = population.row(bestFitIndex); for (int i = 0; i < POPULATION_SIZE; i++) { // crossover crossover(i, population.row(i), bestIndv); } return; } void EvolutionaryAlgorithm::tournament() { std::cout << "TOURNAMENT\n"; arma::mat::fixed<POPULATION_SIZE, NB_PARAMETERS> oldPopulation; int parentIndex[2]; // copying last population (new one will be different) // for (int i = 0; i < POPULATION_SIZE; i++) { // for (int j = 0; j < NB_PARAMETERS; j++) { // oldPopulation[i][j] = population[i][j]; // } // } oldPopulation = population; for (int i = 0; i < POPULATION_SIZE; i++) { if (i == bestFitIndex) continue; // chossing parents for new individual for (int j = 0; j < 2; j++) { int indexIndA = rand() % POPULATION_SIZE; // indv 1 that will "fight" to be parent int indexIndB = rand() % POPULATION_SIZE; // indv 2 that will "fight" to be parent parentIndex[j] = (fitness[indexIndA] < fitness[indexIndB] ? indexIndA : indexIndB); } // crossover crossover(i, oldPopulation.row(parentIndex[0]), oldPopulation.row(parentIndex[1])); } return; } //TODO: swap priorities. Smaller fitness must be morre relevant void EvolutionaryAlgorithm::roulette() { std::cout << "ROULETTE\n"; arma::mat::fixed<POPULATION_SIZE, NB_PARAMETERS> oldPopulation; double standardizedFitness[POPULATION_SIZE]; int parentIndex[2], rNb; double probSum = 0.0, partialSum = 0.0; // copying last population (new one will be different) // for (int i = 0; i < POPULATION_SIZE; i++) { // for (int j = 0; j < NB_PARAMETERS; j++) { // oldPopulation[i][j] = population[i][j]; // } // } oldPopulation = population; // Standardize fitness (set probabilites that add up to 100%) for (int i = 0; i < POPULATION_SIZE; i++) probSum += fitness[i]; for (int i = 0; i < POPULATION_SIZE; i++) standardizedFitness[i] = fitness[i]/probSum; // Chosing new parents for each individual for (int i = 0; i < POPULATION_SIZE; i++) { if (i == bestFitIndex) // preserves best individual continue; for (int k = 0; k < 2; k++) { // chosing 2 parents rNb = ((double) rand() / RAND_MAX); // rand between 0 and 1 partialSum = 0.0; for (int j = 0; j < POPULATION_SIZE; j++) { // randomly chosing and individual according to its fitness (+fitness = +probabity) partialSum += fitness[j]; if (partialSum >= rNb) { parentIndex[k] = j; // new parent at index j break; } } } // crossover crossover(i, oldPopulation.row(parentIndex[0]), oldPopulation.row(parentIndex[1])); } } void EvolutionaryAlgorithm::mutate(int indIndex) { int plusMinusSign = (rand() % 2 ? MUTATE_POSITIVE_MULT : MUTATE_NEGATIVE_MULT); // mutation will increase the param value or decrease it int mutateParamIndex = rand() % NB_PARAMETERS; // index of parameter that will be mutated population(indIndex, mutateParamIndex) += population(indIndex, mutateParamIndex) * mutationRate * plusMinusSign; return; } // 3rd step: selection + mutation and crossover void EvolutionaryAlgorithm::selectionAndMutation() { // tournament, elitism, roulette... std::cout << "SELECTION\n"; // Selection method (+ crossover) (this->*(selectionType[selectionMethod]))(); // void (*selectionType[])() => tournament(), elitism() or roulette() // Mutation // Mutating all params from individuals // arma::rowvec bestIndv = population.row(maxFitIndex); // arma::mat mutateMatrix(POPULATION_SIZE, NB_PARAMETERS, arma::fill::randu); // mutateMatrix = ((mutateMatrix * MAX_PARAM_VALUE) - MAX_PARAM_VALUE/2.0) * mutationRate; // population = population + mutateMatrix; // population.row(maxFitIndex) = bestIndv; // population.transform( [](double x) { return ((x < 0 || x > MAX_PARAM_VALUE) ? abs(MAX_PARAM_VALUE-abs(x)) : x); } ); // Mutating only one parameter from each individual for(int i = 0; i < bestFitIndex; i++) mutate(i); // (don't mutate best one) for(int i = bestFitIndex+1; i < POPULATION_SIZE; i++) mutate(i); return; } // Initialize .csv file to sabe data from the EA // Creates the column's headers // generation, paramsIndv1, fitnessIndv1 ..., paramsIndvN, fitnessIndvN, bestParams, bestFitness void EvolutionaryAlgorithm::createCSV(std::string pathPos) { std::ofstream csvFileWriter; csvFileWriter.open("historyEA-" + pathPos + ".csv"); if (!csvFileWriter.good()) { std::cout << "[!] Error occurred while trying to create historyEA" << pathPos << ".csv!\n"; return; } csvFileWriter << "generation,"; for (int i = 0; i < POPULATION_SIZE; i++) csvFileWriter << "paramsIndv" << i << ",fitnessIndv" << i << ","; csvFileWriter << "paramsBestIndv,fitnessBestIndv\n"; return; } std::string EvolutionaryAlgorithm::formatParamsString(int indvIndex) { std::string paramsFormated = "[ "; for (int i = 0; i < NB_PARAMETERS; i++) { paramsFormated += std::to_string(population(indvIndex, i)); paramsFormated += " "; } paramsFormated += "]"; return paramsFormated; } // Saves information about a generation in a .csv file // Information: generation, paramsIndv1, fitnessIndv1 ..., paramsIndvN, fitnessIndvN, bestParams, bestFitness void EvolutionaryAlgorithm::saveGenerationData(int generation) { std::ofstream csvFileWriter; csvFileWriter.open("historyEA.csv", std::ios_base::app); // append instead of overwrite if (!csvFileWriter.good()) { std::cout << "[!] Error occurred while trying to open historyEA.csv!\n"; return; } csvFileWriter << generation << ","; for (int i = 0; i < POPULATION_SIZE; i++) { csvFileWriter << formatParamsString(i) << "," << fitness[i] << ","; } csvFileWriter << formatParamsString(bestFitIndex) << "," << bestFitness << "\n"; } void EvolutionaryAlgorithm::increaseMutation() { if(!mutationRate) { mutationRate = INITIAL_MUTATION; } else { mutationRate *= MUTATION_INCREASE_RATIO; if (mutationRate > MAX_MUTATION_RATE) // mutation reached its max value mutationRate = MAX_MUTATION_RATE; } return; } // Function that will kill the worst individual each "APPLY_PREDATION_INTERVAL" number of generations void EvolutionaryAlgorithm::predationOfOne() { arma::rowvec newInd(NB_PARAMETERS, arma::fill::randu); // creating totally new individual newInd = newInd * MAX_PARAM_VALUE; population.row(worstFitIndex) = newInd; return; } // A sequence of selections will run, this method will indicate the selection method that will be used by the next batch of generations void EvolutionaryAlgorithm::partIncrease(){ selectionMethod = partsSelectionMethods[++partPos]; return; } // Function that will kill all individuals but the best to reset the population and generate new individuals (without biases) void EvolutionaryAlgorithm::oneRemainingPopReset() { std::cout << "APPLYING POPULATION RESET\n"; arma::rowvec best = population.row(bestFitIndex); initializePop(INITIALS); population.row(0) = best; partIncrease(); return; } // Ends the batch of generations from current EA void EvolutionaryAlgorithm::endEABatch() { std::cout << "END EA BATCH"; continueBatch = false; return; } // Generic function that checks if certain event should happen (predation, population reset, mutation increase...) bool EvolutionaryAlgorithm::eventHappens(int eventType) { if(nbNoImprovementGens % eventTriggerModule[eventType] == 0) return true; // the event being verified should happen! return false; } void EvolutionaryAlgorithm::checkEvents() { if(nbNoImprovementGens == 0) return; for(int i = 5; i > 0; i++){ if(eventHappens(i)) { (this->*(eventTypes[i-1]))(); // calls one of these functions: increaseMutation, predationOfOne, partIncrease, oneRemainingPopReset, fullPopReset return; } } } void EvolutionaryAlgorithm::startEventTriggerList(){ for(int i = 1; i < 6; i++) eventTriggerModule[i] *= eventTriggerModule[i-1]; } void EvolutionaryAlgorithm::evoAlg(std::string csvStr){ nbNoImprovementGens = 0; createCSV(csvStr); continueBatch = true; partPos = 0; // defines selection method of current batch of generations generationIndex = 0; while (continueBatch) { std::cout << "\n==== Generation " << generationIndex << " ====\n"; // population.print("Current population:"); evaluatePop(); selectionAndMutation(); saveGenerationData(generationIndex); checkEvents(); // checks if mutation should increase, predation or population reset should occur etc. // if fullPopReset() is called, continueEA = false generationIndex++; scanf("%*c"); } return; } void EvolutionaryAlgorithm::fillInitialsWithBests(int tournmentType){ for(int i = 0; i < TOURNMENT_RATE; i++) population.row(i) = bestTournmentIndv[tournmentType].row(i); } void EvolutionaryAlgorithm::semiFinalsTournment(int semiFinalPos){ for(int i = 0; i < TOURNMENT_RATE; i++){ initializePop(INITIALS); evoAlg("SF-" + std::to_string(semiFinalPos) + "_EA-" + std::to_string(i)); bestTournmentIndv[SEMI_FINALS].row(i) = population.row(bestFitIndex); // saves the best individual from current EA } initializePop(SEMI_FINALS); evoAlg("SF-" + std::to_string(semiFinalPos)); // this EA will use the best individuals from each previous EA bestTournmentIndv[FINALS].row(semiFinalPos) = population.row(bestFitIndex); // saves the best individual from current EA return; } void EvolutionaryAlgorithm::finalTournment(){ for(int i = 0; i < TOURNMENT_RATE; i++) semiFinalsTournment(i); initializePop(FINALS); evoAlg("Main"); // this EA will use the best individuals from each semifinal bestIndividual = population.row(bestFitIndex); std::cout << "EVOLUTIONARY ALGORITHM FINISHED!" << std::endl; //eaFinished = true; return; } EvolutionaryAlgorithm::EvolutionaryAlgorithm():eaFinished(false){ srand(time(NULL)); startEventTriggerList(); system("clear"); pthread_create(&scriptThread, NULL, runScript, this); } EvolutionaryAlgorithm::~EvolutionaryAlgorithm(){ pthread_join(scriptThread, NULL); } void *runScript(void *scriptFuncObject) { EvolutionaryAlgorithm *script = (EvolutionaryAlgorithm*) scriptFuncObject; script->finalTournment(); return NULL; } void EvolutionaryAlgorithm::setFitness(int pos,std::vector<std::pair<int,int>> fitnessResults){ fitness[pos] = 0; for(auto fitnessPair : fitnessResults){ fitness[pos] += std::abs(fitnessPair.first - fitnessPair.second)/fitnessPair.first; } remainingFitnessToCalc -= 1; return; }
{ "alphanum_fraction": 0.6446322379, "author": null, "avg_line_length": 34.0618336887, "converted": null, "ext": "cpp", "file": null, "hexsha": "817f3f3e140d414af2ed2a806bf75ed146ffc175", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "2f694ec2a36969af51162556b40fd513d6c1213e", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "Haltz01/GA-AdjustmentGameParameters", "max_forks_repo_path": "EvolutionaryAlgorithm.cpp", "max_issues_count": null, "max_issues_repo_head_hexsha": "2f694ec2a36969af51162556b40fd513d6c1213e", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "Haltz01/GA-AdjustmentGameParameters", "max_issues_repo_path": "EvolutionaryAlgorithm.cpp", "max_line_length": 168, "max_stars_count": null, "max_stars_repo_head_hexsha": "2f694ec2a36969af51162556b40fd513d6c1213e", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "Haltz01/GA-AdjustmentGameParameters", "max_stars_repo_path": "EvolutionaryAlgorithm.cpp", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 4083, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 15975 }
abstract type distribution end module genericDistribution <: distribution #default values E(x)=0 #mean value #1st moment generating function σ(x)=1 #variance #2nd moment generating function function getPdf(x=smoothFunction) """ generalization of (any ) """ return pdf(x)= smoothFunction end function pdf(x, range) return pdf(x, range) function cdf(x, range) #TODO: approx Integral of pdf(x) return sum(pdf(x,range)) #sum of all points #assumes all points are reachable, and funcion is continuoous on all of them #no-Jumps model end # module genericDistribution() genericDistribution()
{ "alphanum_fraction": 0.7398373984, "author": null, "avg_line_length": 26.7391304348, "converted": null, "ext": "jl", "file": null, "hexsha": "a7f53b14cd62701c96c96999ba0ec9d89f90e6bb", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "f920e93af5ece5c4f8751bf493047770689ecaff", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "adamwillisXanax/DeepLearner", "max_forks_repo_path": "src/Distributions/genericDistribution.jl", "max_issues_count": 1, "max_issues_repo_head_hexsha": "f920e93af5ece5c4f8751bf493047770689ecaff", "max_issues_repo_issues_event_max_datetime": "2022-02-01T05:33:04.000Z", "max_issues_repo_issues_event_min_datetime": "2022-01-23T12:20:47.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "adamwillisMastery/DeepLearner", "max_issues_repo_path": "src/Distributions/genericDistribution.jl", "max_line_length": 137, "max_stars_count": 2, "max_stars_repo_head_hexsha": "f920e93af5ece5c4f8751bf493047770689ecaff", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "adamwillisMastery/DeepLearner", "max_stars_repo_path": "src/Distributions/genericDistribution.jl", "max_stars_repo_stars_event_max_datetime": "2022-01-17T12:38:14.000Z", "max_stars_repo_stars_event_min_datetime": "2022-01-15T00:43:05.000Z", "num_tokens": 148, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 615 }
from collections import OrderedDict import copy import cPickle import gzip import os import urllib import random import stat import subprocess import sys import time import numpy import theano from theano import tensor as T PREFIX = os.getenv( 'ATISDATA', os.path.join(os.path.split(os.path.abspath(os.path.dirname(__file__)))[0], 'data')) # utils functions def shuffle(lol, seed): ''' lol :: list of list as input seed :: seed the shuffling shuffle inplace each list in the same order ''' for l in lol: random.seed(seed) random.shuffle(l) # start-snippet-1 def contextwin(l, win): ''' win :: int corresponding to the size of the window given a list of indexes composing a sentence l :: array containing the word indexes it will return a list of list of indexes corresponding to context windows surrounding each word in the sentence ''' assert (win % 2) == 1 assert win >= 1 l = list(l) lpadded = win // 2 * [-1] + l + win // 2 * [-1] out = [lpadded[i:(i + win)] for i in range(len(l))] assert len(out) == len(l) return out # end-snippet-1 # data loading functions def atisfold(fold): assert fold in range(5) filename = os.path.join(PREFIX, 'atis.fold'+str(fold)+'.pkl.gz') f = gzip.open(filename, 'rb') train_set, valid_set, test_set, dicts = cPickle.load(f) return train_set, valid_set, test_set, dicts # metrics function using conlleval.pl def conlleval(p, g, w, filename, script_path): ''' INPUT: p :: predictions g :: groundtruth w :: corresponding words OUTPUT: filename :: name of the file where the predictions are written. it will be the input of conlleval.pl script for computing the performance in terms of precision recall and f1 score OTHER: script_path :: path to the directory containing the conlleval.pl script ''' out = '' for sl, sp, sw in zip(g, p, w): out += 'BOS O O\n' for wl, wp, w in zip(sl, sp, sw): out += w + ' ' + wl + ' ' + wp + '\n' out += 'EOS O O\n\n' f = open(filename, 'w') f.writelines(out) f.close() return get_perf(filename, script_path) def download(origin, destination): ''' download the corresponding atis file from http://www-etud.iro.umontreal.ca/~mesnilgr/atis/ ''' print 'Downloading data from %s' % origin urllib.urlretrieve(origin, destination) def get_perf(filename, folder): ''' run conlleval.pl perl script to obtain precision/recall and F1 score ''' _conlleval = os.path.join(folder, 'conlleval.pl') if not os.path.isfile(_conlleval): url = 'http://www-etud.iro.umontreal.ca/~mesnilgr/atis/conlleval.pl' download(url, _conlleval) os.chmod(_conlleval, stat.S_IRWXU) # give the execute permissions proc = subprocess.Popen(["perl", _conlleval], stdin=subprocess.PIPE, stdout=subprocess.PIPE) stdout, _ = proc.communicate(''.join(open(filename).readlines())) for line in stdout.split('\n'): if 'accuracy' in line: out = line.split() break precision = float(out[6][:-2]) recall = float(out[8][:-2]) f1score = float(out[10]) return {'p': precision, 'r': recall, 'f1': f1score} # start-snippet-2 class RNNSLU(object): ''' elman neural net model ''' def __init__(self, nh, nc, ne, de, cs): ''' nh :: dimension of the hidden layer nc :: number of classes ne :: number of word embeddings in the vocabulary de :: dimension of the word embeddings cs :: word window context size ''' # parameters of the model self.emb = theano.shared(name='embeddings', value=0.2 * numpy.random.uniform(-1.0, 1.0, (ne+1, de)) # add one for padding at the end .astype(theano.config.floatX)) self.wx = theano.shared(name='wx', value=0.2 * numpy.random.uniform(-1.0, 1.0, (de * cs, nh)) .astype(theano.config.floatX)) self.wh = theano.shared(name='wh', value=0.2 * numpy.random.uniform(-1.0, 1.0, (nh, nh)) .astype(theano.config.floatX)) self.w = theano.shared(name='w', value=0.2 * numpy.random.uniform(-1.0, 1.0, (nh, nc)) .astype(theano.config.floatX)) self.bh = theano.shared(name='bh', value=numpy.zeros(nh, dtype=theano.config.floatX)) self.b = theano.shared(name='b', value=numpy.zeros(nc, dtype=theano.config.floatX)) self.h0 = theano.shared(name='h0', value=numpy.zeros(nh, dtype=theano.config.floatX)) # bundle self.params = [self.emb, self.wx, self.wh, self.w, self.bh, self.b, self.h0] # end-snippet-2 # as many columns as context window size # as many lines as words in the sentence # start-snippet-3 idxs = T.imatrix() x = self.emb[idxs].reshape((idxs.shape[0], de*cs)) y_sentence = T.ivector('y_sentence') # labels # end-snippet-3 start-snippet-4 def recurrence(x_t, h_tm1): h_t = T.nnet.sigmoid(T.dot(x_t, self.wx) + T.dot(h_tm1, self.wh) + self.bh) s_t = T.nnet.softmax(T.dot(h_t, self.w) + self.b) return [h_t, s_t] [h, s], _ = theano.scan(fn=recurrence, sequences=x, outputs_info=[self.h0, None], n_steps=x.shape[0]) p_y_given_x_sentence = s[:, 0, :] y_pred = T.argmax(p_y_given_x_sentence, axis=1) # end-snippet-4 # cost and gradients and learning rate # start-snippet-5 lr = T.scalar('lr') sentence_nll = -T.mean(T.log(p_y_given_x_sentence) [T.arange(x.shape[0]), y_sentence]) sentence_gradients = T.grad(sentence_nll, self.params) sentence_updates = OrderedDict((p, p - lr*g) for p, g in zip(self.params, sentence_gradients)) # end-snippet-5 # theano functions to compile # start-snippet-6 self.classify = theano.function(inputs=[idxs], outputs=y_pred) self.sentence_train = theano.function(inputs=[idxs, y_sentence, lr], outputs=sentence_nll, updates=sentence_updates) # end-snippet-6 start-snippet-7 self.normalize = theano.function(inputs=[], updates={self.emb: self.emb / T.sqrt((self.emb**2) .sum(axis=1)) .dimshuffle(0, 'x')}) # end-snippet-7 def train(self, x, y, window_size, learning_rate): cwords = contextwin(x, window_size) words = map(lambda x: numpy.asarray(x).astype('int32'), cwords) labels = y self.sentence_train(words, labels, learning_rate) self.normalize() def save(self, folder): for param in self.params: numpy.save(os.path.join(folder, param.name + '.npy'), param.get_value()) def load(self, folder): for param in self.params: param.set_value(numpy.load(os.path.join(folder, param.name + '.npy'))) def main(param=None): if not param: param = { 'fold': 3, # 5 folds 0,1,2,3,4 'data': 'atis', 'lr': 0.0970806646812754, 'verbose': 1, 'decay': True, # decay on the learning rate if improvement stops 'win': 7, # number of words in the context window 'nhidden': 200, # number of hidden units 'seed': 345, 'emb_dimension': 50, # dimension of word embedding 'nepochs': 60, # 60 is recommended 'savemodel': False} print param folder_name = os.path.basename(__file__).split('.')[0] folder = os.path.join(os.path.dirname(__file__), folder_name) if not os.path.exists(folder): os.mkdir(folder) # load the dataset train_set, valid_set, test_set, dic = atisfold(param['fold']) idx2label = dict((k, v) for v, k in dic['labels2idx'].iteritems()) idx2word = dict((k, v) for v, k in dic['words2idx'].iteritems()) train_lex, train_ne, train_y = train_set valid_lex, valid_ne, valid_y = valid_set test_lex, test_ne, test_y = test_set vocsize = len(set(reduce(lambda x, y: list(x) + list(y), train_lex + valid_lex + test_lex))) nclasses = len(set(reduce(lambda x, y: list(x)+list(y), train_y + test_y + valid_y))) nsentences = len(train_lex) groundtruth_valid = [map(lambda x: idx2label[x], y) for y in valid_y] words_valid = [map(lambda x: idx2word[x], w) for w in valid_lex] groundtruth_test = [map(lambda x: idx2label[x], y) for y in test_y] words_test = [map(lambda x: idx2word[x], w) for w in test_lex] # instanciate the model numpy.random.seed(param['seed']) random.seed(param['seed']) rnn = RNNSLU(nh=param['nhidden'], nc=nclasses, ne=vocsize, de=param['emb_dimension'], cs=param['win']) # train with early stopping on validation set best_f1 = -numpy.inf param['clr'] = param['lr'] for e in xrange(param['nepochs']): # shuffle shuffle([train_lex, train_ne, train_y], param['seed']) param['ce'] = e tic = time.time() for i, (x, y) in enumerate(zip(train_lex, train_y)): rnn.train(x, y, param['win'], param['clr']) print '[learning] epoch %i >> %2.2f%%' % ( e, (i + 1) * 100. / nsentences), print 'completed in %.2f (sec) <<\r' % (time.time() - tic), sys.stdout.flush() # evaluation // back into the real world : idx -> words predictions_test = [map(lambda x: idx2label[x], rnn.classify(numpy.asarray( contextwin(x, param['win'])).astype('int32'))) for x in test_lex] predictions_valid = [map(lambda x: idx2label[x], rnn.classify(numpy.asarray( contextwin(x, param['win'])).astype('int32'))) for x in valid_lex] # evaluation // compute the accuracy using conlleval.pl res_test = conlleval(predictions_test, groundtruth_test, words_test, folder + '/current.test.txt', folder) res_valid = conlleval(predictions_valid, groundtruth_valid, words_valid, folder + '/current.valid.txt', folder) if res_valid['f1'] > best_f1: if param['savemodel']: rnn.save(folder) best_rnn = copy.deepcopy(rnn) best_f1 = res_valid['f1'] if param['verbose']: print('NEW BEST: epoch', e, 'valid F1', res_valid['f1'], 'best test F1', res_test['f1']) param['vf1'], param['tf1'] = res_valid['f1'], res_test['f1'] param['vp'], param['tp'] = res_valid['p'], res_test['p'] param['vr'], param['tr'] = res_valid['r'], res_test['r'] param['be'] = e subprocess.call(['mv', folder + '/current.test.txt', folder + '/best.test.txt']) subprocess.call(['mv', folder + '/current.valid.txt', folder + '/best.valid.txt']) else: if param['verbose']: print '' # learning rate decay if no improvement in 10 epochs if param['decay'] and abs(param['be']-param['ce']) >= 10: param['clr'] *= 0.5 rnn = best_rnn if param['clr'] < 1e-5: break print('BEST RESULT: epoch', param['be'], 'valid F1', param['vf1'], 'best test F1', param['tf1'], 'with the model', folder) if __name__ == '__main__': main()
{ "alphanum_fraction": 0.5116349123, "author": null, "avg_line_length": 34.0487179487, "converted": null, "ext": "py", "file": null, "hexsha": "4ab80b54b27e6bf293f12223a4a70e109b8a6252", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "d2d1b0148989187c1433597f9c3ae4357178c082", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "fegonda/icon_demo", "max_forks_repo_path": "code/external/rnnslu.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "d2d1b0148989187c1433597f9c3ae4357178c082", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "fegonda/icon_demo", "max_issues_repo_path": "code/external/rnnslu.py", "max_line_length": 78, "max_stars_count": null, "max_stars_repo_head_hexsha": "d2d1b0148989187c1433597f9c3ae4357178c082", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "fegonda/icon_demo", "max_stars_repo_path": "code/external/rnnslu.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 3121, "path": null, "reason": "import numpy,import theano,from theano", "repo": null, "save_path": null, "sha": null, "size": 13279 }
# Copyright 2021 Huawei Technologies Co., Ltd # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # ============================================================================ """export checkpoint file into air, onnx, mindir models""" import argparse import ast import os import numpy as np from mindspore import Tensor, context, load_checkpoint, load_param_into_net, export import mindspore.common.dtype as mstype from mindspore import nn from src.cell import WithLossCellD, WithLossCellG from src.dcgan import DCGAN from src.discriminator import Discriminator from src.generator import Generator from src.config import dcgan_imagenet_cfg as cfg parser = argparse.ArgumentParser(description='ntsnet export') parser.add_argument("--run_modelart", type=ast.literal_eval, default=False, help="Run on modelArt, default is false.") parser.add_argument("--device_id", type=int, default=0, help="Device id") parser.add_argument("--batch_size", type=int, default=128, help="batch size") parser.add_argument("--ckpt_file", type=str, required=True, help="Checkpoint file name.") parser.add_argument('--data_url', default=None, help='Directory contains CUB_200_2011 dataset.') parser.add_argument('--train_url', default=None, help='Directory contains checkpoint file') parser.add_argument("--file_name", type=str, default="ntsnet", help="output file name.") parser.add_argument("--file_format", type=str, default="MINDIR", help="file format") parser.add_argument('--device_target', type=str, default="Ascend", choices=['Ascend', 'GPU', 'CPU'], help='device target (default: Ascend)') args = parser.parse_args() context.set_context(mode=context.GRAPH_MODE, device_target=args.device_target) if args.device_target == "Ascend": context.set_context(device_id=args.device_id) if __name__ == '__main__': netD = Discriminator() netG = Generator() criterion = nn.BCELoss(reduction='mean') netD_with_criterion = WithLossCellD(netD, netG, criterion) netG_with_criterion = WithLossCellG(netD, netG, criterion) optimizerD = nn.Adam(netD.trainable_params(), learning_rate=cfg.learning_rate, beta1=cfg.beta1) optimizerG = nn.Adam(netG.trainable_params(), learning_rate=cfg.learning_rate, beta1=cfg.beta1) myTrainOneStepCellForD = nn.TrainOneStepCell(netD_with_criterion, optimizerD) myTrainOneStepCellForG = nn.TrainOneStepCell(netG_with_criterion, optimizerG) net = DCGAN(myTrainOneStepCellForD, myTrainOneStepCellForG) param_dict = load_checkpoint(os.path.join(args.train_url, args.ckpt_file)) load_param_into_net(net, param_dict) net.set_train(False) # inputs = Tensor(np.random.rand(args.batch_size, 3, 448, 448), mstype.float32) real_data = Tensor(np.random.rand(args.batch_size, 3, 32, 32), mstype.float32) latent_code = Tensor(np.random.rand(args.batch_size, 100, 1, 1), mstype.float32) inputs = [real_data, latent_code] export(net, *inputs, file_name=args.file_name, file_format=args.file_format)
{ "alphanum_fraction": 0.74734119, "author": null, "avg_line_length": 47.6575342466, "converted": null, "ext": "py", "file": null, "hexsha": "a5d53fbbbde0aaca332183cb7f46700e5f751e6f", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "1331c7e432fb691d1cfa625ab7cc7451dcfc7ce0", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "LottieWang/mindspore", "max_forks_repo_path": "model_zoo/research/cv/dcgan/export.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "1331c7e432fb691d1cfa625ab7cc7451dcfc7ce0", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "LottieWang/mindspore", "max_issues_repo_path": "model_zoo/research/cv/dcgan/export.py", "max_line_length": 118, "max_stars_count": null, "max_stars_repo_head_hexsha": "1331c7e432fb691d1cfa625ab7cc7451dcfc7ce0", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "LottieWang/mindspore", "max_stars_repo_path": "model_zoo/research/cv/dcgan/export.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 837, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 3479 }
# influence functions for shapley values def shapley_influence_function(Z, z_counts, W, v, psi, G, c_n, ics, measure): """ Compute influence function for the given predictiveness measure @param Z the subsets of the power set with estimates @param W the matrix of weights @param v the estimated predictivness @param psi the estimated Shapley values @param G the constrained ls matrix @param c_n the constraints @param ics a list of all ics @param measure the predictiveness measure """ import numpy as np ## compute contribution from estimating V Z_W = Z.transpose().dot(W) A_m = Z_W.dot(Z) A_m_inv = np.linalg.inv(A_m) phi_01 = A_m_inv.dot(Z_W).dot(ics) ## compute contribution from estimating Q qr_decomp = np.linalg.qr(G.transpose(), mode = 'complete') U_2 = qr_decomp[0][:, 3:(Z.shape[1])] V = U_2.transpose().dot(Z.transpose().dot(W).dot(Z)).dot(U_2) phi_02_shared_mat = (-1) * U_2.dot(np.linalg.inv(V)) phi_02_uniq_vectors = np.array([(Z[z, :].dot(psi) - v[z]) * (U_2.transpose().dot(Z[z, :])) for z in range(Z.shape[0])], dtype = np.float64).transpose() phi_02_uniq = phi_02_shared_mat.dot(phi_02_uniq_vectors) phi_02 = np.repeat(phi_02_uniq, z_counts, axis=1) return {'contrib_v': phi_01, 'contrib_s': phi_02} def shapley_se(shapley_ics, idx, gamma, na_rm = True): """ Standard error for the desired Shapley value @param shapley_ics: all influence function estimates @param idx: the index of interest @param gamma: the constant for sampling @param na_rm: remove NaNs? @return the standard error corresponding to the shapley value at idx """ import numpy as np if na_rm: var_v = np.nanvar(shapley_ics['contrib_v'][idx, :]) var_s = np.nanvar(shapley_ics['contrib_s'][idx, :]) else: var_v = np.var(shapley_ics['contrib_v'][idx, :]) var_s = np.var(shapley_ics['contrib_s'][idx, :]) se = np.sqrt(var_v / shapley_ics['contrib_v'].shape[1] + var_s / shapley_ics['contrib_s'].shape[1] / gamma) return se
{ "alphanum_fraction": 0.6676190476, "author": null, "avg_line_length": 37.5, "converted": null, "ext": "py", "file": null, "hexsha": "b41c723de4428c8f8fa5b297281a5294734406c4", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "681eb21e1ff1141dc9fbaa35261e24dd17296857", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "bdwilliamson/npvipy", "max_forks_repo_path": "vimpy/spvim_ic.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "681eb21e1ff1141dc9fbaa35261e24dd17296857", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "bdwilliamson/npvipy", "max_issues_repo_path": "vimpy/spvim_ic.py", "max_line_length": 155, "max_stars_count": null, "max_stars_repo_head_hexsha": "681eb21e1ff1141dc9fbaa35261e24dd17296857", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "bdwilliamson/npvipy", "max_stars_repo_path": "vimpy/spvim_ic.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 626, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2100 }
import numpy as np from pengle.transformer.base import FeatureOverwriter, timer import copy class ComplementMissingValue(FeatureOverwriter): def __init__(self, columns, agg_func=np.mean): super().__init__(columns) self.agg_func = agg_func def apply(self, train_dataset, test_dataset): train = copy.deepcopy(train_dataset) test = copy.deepcopy(test_dataset) for column in self.columns: agg_result = self.agg_func(train_dataset.data[column]) train.data[column].fillna(agg_result, inplace=True) test.data[column].fillna(agg_result, inplace=True) return train, test class ExtractStrings(FeatureOverwriter): def __init__(self, columns, regexps): super().__init__(columns) self.regexps = regexps def apply(self, train_dataset, test_dataset): train = copy.deepcopy(train_dataset) test = copy.deepcopy(test_dataset) for column, regexp in zip(self.columns, self.regexps): train.data[column] = train_dataset.data[column].str.extract( regexp, expand=False) test.data[column] = test_dataset.data[column].str.extract( regexp, expand=False) return train, test class ReplaceStrings(FeatureOverwriter): def __init__(self, columns, replace_rule): super().__init__(columns) self.replace_rule = replace_rule def apply(self, train_dataset, test_dataset): train = copy.deepcopy(train_dataset) test = copy.deepcopy(test_dataset) for column in self.columns: train.data[column].replace(self.replace_rule, inplace=True) test.data[column].replace(self.replace_rule, inplace=True) return train, test
{ "alphanum_fraction": 0.6727066818, "author": null, "avg_line_length": 36.0408163265, "converted": null, "ext": "py", "file": null, "hexsha": "406874c7e5050d3199d65f5e4d48e83ed2507c8c", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "cd8c60f953ee7936bf25af84503fac8092ef7b94", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "pesuchin/pengle", "max_forks_repo_path": "pengle/transformer/preprocessor.py", "max_issues_count": 8, "max_issues_repo_head_hexsha": "cd8c60f953ee7936bf25af84503fac8092ef7b94", "max_issues_repo_issues_event_max_datetime": "2020-05-30T02:19:49.000Z", "max_issues_repo_issues_event_min_datetime": "2020-05-16T05:44:57.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "pesuchin/pengle", "max_issues_repo_path": "pengle/transformer/preprocessor.py", "max_line_length": 72, "max_stars_count": null, "max_stars_repo_head_hexsha": "cd8c60f953ee7936bf25af84503fac8092ef7b94", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "pesuchin/pengle", "max_stars_repo_path": "pengle/transformer/preprocessor.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 363, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1766 }
import chainer import chainer.functions as F import chainer.links as L import onnx import onnx.helper as oh from onnx import numpy_helper from onnx import TensorProto from onnx import ModelProto from chainer_compiler.elichika.parser import core from chainer_compiler.elichika.parser import graphs from chainer_compiler.elichika.parser import values from chainer_compiler.elichika.parser import nodes from chainer_compiler.elichika.parser import functions from chainer_compiler.elichika.parser import functions_builtin from chainer_compiler.elichika.parser import utils import numpy as np import collections from chainer_compiler.elichika import onnx_converters as oc def _pair(x): if isinstance(x, collections.Iterable): return x return (x, x) def _list(v) -> 'List[int]': if isinstance(v, collections.Iterable): return list(x for x in v) return [v] def get_onnx_dtype(dtype): a = np.zeros((), dtype=dtype) dt = onnx.mapping.NP_TYPE_TO_TENSOR_TYPE[a.dtype] return dt class BaseConverter(object): def __init__(self): self.expected_args = () def parse_args(self, onnx_graph, node): assert hasattr( self, 'expected_args'), 'BaseConverter subclass must have `expected_args`' parser = oc.NodeParse() for arg_def in self.expected_args: parser.add_def(*arg_def) parser.parse(onnx_graph, node) return parser def __call__(self, onnx_graph, node): raise NotImplementedError class ConverterRelu(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In),) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'Relu', [parser.get('x')], node.outputs, name=str(node.lineprop)) class ConverterElu(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In), ('alpha', oc.ParseType.Att)) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'Elu', [parser.get('x')], node.outputs, name=str(node.lineprop), alpha=parser.get('alpha')) class ConverterLeakyRelu(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In), ('slope', oc.ParseType.Att)) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'LeakyRelu', [parser.get('x')], node.outputs, name=str(node.lineprop), alpha=parser.get('slope')) class ConverterLogSoftmax(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In), ('axis', oc.ParseType.Att)) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'LogSoftmax', [parser.get('x')], node.outputs, name=str(node.lineprop), axis=parser.get('axis'), chainer_is_onnx_semantics=False) class ConverterSelu(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In), ('alpha', oc.ParseType.Att), ('scale', oc.ParseType.Att)) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'Selu', [parser.get('x')], node.outputs, name=str(node.lineprop), alpha=parser.get('alpha'), gamma=parser.get('scale'), ) class ConverterSigmoid(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In),) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'Sigmoid', [parser.get('x')], node.outputs, name=str(node.lineprop)) class ConverterSoftmax(BaseConverter): def __init__(self): self.expected_args = ( ('x', oc.ParseType.In), ('axis', oc.ParseType.Att)) def __call__(self, onnx_graph, node): parser = self.parse_args(onnx_graph, node) onnx_graph.add_node( 'Softmax', [parser.get('x')], node.outputs, name=str(node.lineprop), axis=parser.get('axis'), chainer_is_onnx_semantics=False)
{ "alphanum_fraction": 0.6059701493, "author": null, "avg_line_length": 26.4971751412, "converted": null, "ext": "py", "file": null, "hexsha": "f2b898e02c1aa7075b806196125c0a2caf17aa97", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 26, "max_forks_repo_forks_event_max_datetime": "2021-11-26T04:24:35.000Z", "max_forks_repo_forks_event_min_datetime": "2019-01-25T07:21:09.000Z", "max_forks_repo_head_hexsha": "5f5ad365d14398d6ae0214fa012eb10360db8e7e", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "vermashresth/chainer-compiler", "max_forks_repo_path": "chainer_compiler/elichika/functions_chainer_activation.py", "max_issues_count": 431, "max_issues_repo_head_hexsha": "5f5ad365d14398d6ae0214fa012eb10360db8e7e", "max_issues_repo_issues_event_max_datetime": "2020-06-17T05:28:55.000Z", "max_issues_repo_issues_event_min_datetime": "2019-01-25T10:18:44.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "vermashresth/chainer-compiler", "max_issues_repo_path": "chainer_compiler/elichika/functions_chainer_activation.py", "max_line_length": 86, "max_stars_count": 116, "max_stars_repo_head_hexsha": "5f5ad365d14398d6ae0214fa012eb10360db8e7e", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "vermashresth/chainer-compiler", "max_stars_repo_path": "chainer_compiler/elichika/functions_chainer_activation.py", "max_stars_repo_stars_event_max_datetime": "2022-03-08T00:11:14.000Z", "max_stars_repo_stars_event_min_datetime": "2019-01-25T03:54:44.000Z", "num_tokens": 1074, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4690 }
import argparse import logging import os import sys import numpy as np import torch import torch.nn as nn import torch.nn.functional as F import yaml from torch import optim from torch.utils.data import DataLoader, random_split from torch.utils.tensorboard import SummaryWriter from tqdm import tqdm from evaluation import * # from unet import U_Net, R2U_Net, AttU_Net, R2AttU_Net, init_weights from models import ChooseModel from utils.dataset import BasicDataset conf = yaml.load(open(os.path.join( sys.path[0], 'config', 'evaluate.yaml')), Loader=yaml.FullLoader) dir_img = conf['DATASET']['IMGS_DIR'] dir_mask = conf['DATASET']['MASKS_DIR'] def count_param(model): param_count = 0 for param in model.parameters(): param_count += param.view(-1).size()[0] return param_count def evaluate_net(net, device, batch_size=16, img_scale=0.5, classes=2): dataset = BasicDataset(dir_img, dir_mask, img_scale, train=True, classes=classes) val_loader = DataLoader(dataset, batch_size=batch_size, shuffle=False, num_workers=8, pin_memory=True) n_val = len(dataset) writer = SummaryWriter( comment=f'BS_{batch_size}_SCALE_{img_scale}') global_step = 0 logging.info(f'''Starting training: Batch size: {batch_size} Device: {device.type} Images scaling: {img_scale} ''') if net.n_classes > 1: criterion = nn.CrossEntropyLoss() else: criterion = nn.BCEWithLogitsLoss() net.eval() epoch_loss = 0 tot = 0 PA = 0 OA = 0 pre = 0 recal = 0 f1s = 0 params = None for batch in tqdm(val_loader): imgs = batch['image'] true_masks = batch['mask'] assert imgs.shape[1] == net.n_channels, \ f'Network has been defined with {net.n_channels} input channels, ' \ f'but loaded images have {imgs.shape[1]} channels. Please check that ' \ 'the images are loaded correctly.' imgs = imgs.to(device=device, dtype=torch.float32) if params is None: params = count_param(net) mask_type = torch.float32 if net.n_classes == 1 else torch.long true_masks = true_masks.to(device=device, dtype=mask_type) if net.n_classes > 1: b, c, w, h = true_masks.shape true_masks = true_masks.view(b, w, h) masks_pred = net(imgs) loss = criterion(masks_pred, true_masks) epoch_loss += loss.item() writer.add_scalar('Loss/train', loss.item(), global_step) for true_mask, pred in zip(true_masks, masks_pred): pred = (pred > 0.5).float() if net.n_classes > 1: tot += F.cross_entropy(pred.unsqueeze(dim=0), true_mask.unsqueeze(dim=0)).item() else: PA += pixel_accuracy(pred, true_mask.squeeze(dim=1)).item() tot += dice_coeff(pred, true_mask.squeeze(dim=1)).item() OA += overall_accuracy(pred, true_mask.squeeze(dim=1)).item() pre += precision(pred, true_mask.squeeze(dim=1)).item() recal += recall(pred, true_mask.squeeze(dim=1)).item() f1s += f1score(pred, true_mask.squeeze(dim=1)).item() epoch_loss /= n_val tot /= n_val PA /= n_val OA /= n_val pre /= n_val recal /= n_val f1s /= n_val if net.n_classes > 1: logging.info(f'Validation loss:{epoch_loss}') logging.info( 'Validation cross entropy: {}'.format(tot)) logging.info(f'Params in this model is: {params}') writer.add_scalar('Loss/test', tot, global_step) else: logging.info(f'Validation loss:{epoch_loss}') logging.info( 'Validation Dice Coeff: {}'.format(tot)) writer.add_scalar('Dice/test', tot, global_step) logging.info( 'Validation Pixel Accuracy: {}'.format(PA)) writer.add_scalar('pA/test', PA, global_step) logging.info( 'Validation Overall Accuracy: {}'.format(OA)) writer.add_scalar('oA/test', OA, global_step) logging.info( 'Validation Precision: {}'.format(pre)) writer.add_scalar('precision/test', pre, global_step) logging.info( 'Validation Recall: {}'.format(recal)) writer.add_scalar('recall/test', recal, global_step) logging.info( 'Validation F1-score: {}'.format(f1s)) writer.add_scalar( 'F1-score/test', f1s, global_step) logging.info(f'Params in this model is: {params}') writer.close() def get_args(): parser = argparse.ArgumentParser(description='Train the UNet on images and target masks', formatter_class=argparse.ArgumentDefaultsHelpFormatter) parser.add_argument('-n', '--network', metavar='NETWORK', type=str, default=conf['MODEL']['MODEL_NAME'], help='network type', dest='network') parser.add_argument('-b', '--batch-size', metavar='B', type=int, nargs='?', default=conf['BATCH_SIZE'], help='Batch size', dest='batchsize') parser.add_argument('-f', '--load', dest='load', type=str, default=conf['MODEL']['LOAD_PATH'], help='Load model from a .pth file') parser.add_argument('-s', '--scale', dest='scale', type=float, default=conf['SCALE'], help='Downscaling factor of the images') return parser.parse_args() if __name__ == '__main__': logging.basicConfig(level=logging.INFO, format='%(levelname)s: %(message)s') args = get_args() device = torch.device('cuda' if torch.cuda.is_available( ) and conf['DEVICE'].lower() == 'cuda' else 'cpu') logging.info(f'Using device {device}') # Change here to adapt to your data # n_channels=3 for RGB images # n_classes is the number of probabilities you want to get per pixel # - For 1 class and background, use n_classes=1 # - For 2 classes, use n_classes=1 # - For N > 2 classes, use n_classes=N network = args.network.lower() net = ChooseModel(network)( n_channels=3, n_classes=conf['DATASET']['NUM_CLASSES']) assert net is not None, f'check your argument --network' # net = AttU_Net(n_channels=3,n_classes=1) # net = AttU_Net(n_channels=3, n_classes=1) logging.info(f'Network:\n' f'\t{net.n_channels} input channels\n' f'\t{net.n_classes} output channels (classes)\n' f'\t{"Bilinear" if net.bilinear else "Dilated conv"} upscaling\n') if args.load: net.load_state_dict( torch.load(args.load, map_location=device) ) logging.info(f'Model loaded from {args.load}') net.to(device=device) # faster convolutions, but more memory # cudnn.benchmark = True try: evaluate_net(net=net, batch_size=args.batchsize, device=device, img_scale=args.scale, classes=conf['DATASET']['NUM_CLASSES']) except KeyboardInterrupt: torch.save(net.state_dict(), 'INTERRUPTED.pth') logging.info('Saved interrupt') try: sys.exit(0) except SystemExit: os._exit(0)
{ "alphanum_fraction": 0.5822586948, "author": null, "avg_line_length": 36.2122641509, "converted": null, "ext": "py", "file": null, "hexsha": "2a3790dca9b3ad7a15a57d4bee446ccfe0cda367", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "7a2c21346739a79c33e7a7ccc081018821868eb7", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "QinchengZhang/PathologySegmentation", "max_forks_repo_path": "Training/pytorch/evaluate.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "7a2c21346739a79c33e7a7ccc081018821868eb7", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "QinchengZhang/PathologySegmentation", "max_issues_repo_path": "Training/pytorch/evaluate.py", "max_line_length": 107, "max_stars_count": 3, "max_stars_repo_head_hexsha": "7a2c21346739a79c33e7a7ccc081018821868eb7", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "QinchengZhang/PathologySegmentation", "max_stars_repo_path": "Training/pytorch/evaluate.py", "max_stars_repo_stars_event_max_datetime": "2021-12-23T01:42:56.000Z", "max_stars_repo_stars_event_min_datetime": "2020-10-25T06:18:21.000Z", "num_tokens": 1751, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 7677 }
""" ============================ StrongTree Fit Example ============================ An example of fitting a StrongTree decision tree using :class:`trees.StrongTree.StrongTreeClassifier` """ from pickle import TRUE import pandas as pd import numpy as np from sklearn.model_selection import train_test_split from trees.StrongTree import StrongTreeClassifier data = pd.read_csv("./data/balance-scale_enc.csv") y = data.pop("target") X_train, X_test, y_train, y_test = train_test_split( data, y, test_size=0.33, random_state=42 ) stcl = StrongTreeClassifier( depth = 1, time_limit = 60, _lambda = 0, benders_oct= False, num_threads=None, obj_mode = 'acc' ) stcl.fit(X_train, y_train, verbose=True) stcl.print_tree() test_pred = stcl.predict(X_test) print('The out-of-sample acc is {}'.format(np.sum(test_pred==y_test)/y_test.shape[0]))
{ "alphanum_fraction": 0.6896551724, "author": null, "avg_line_length": 25.5882352941, "converted": null, "ext": "py", "file": null, "hexsha": "4340e20e891e4239ee975c1eb55d8e739176922a", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "9a921e59cb2578c2a344dd2fa0b88bacb3571b5f", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "nathanaj99/decision_tree_estimators", "max_forks_repo_path": "examples/fit_StrongTree_example.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "9a921e59cb2578c2a344dd2fa0b88bacb3571b5f", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "nathanaj99/decision_tree_estimators", "max_issues_repo_path": "examples/fit_StrongTree_example.py", "max_line_length": 101, "max_stars_count": null, "max_stars_repo_head_hexsha": "9a921e59cb2578c2a344dd2fa0b88bacb3571b5f", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "nathanaj99/decision_tree_estimators", "max_stars_repo_path": "examples/fit_StrongTree_example.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 222, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 870 }
#include <server_lib/emergency_helper.h> #if !defined(STACKTRACE_DISABLED) #include <boost/version.hpp> #include <boost/filesystem.hpp> #include <iostream> #include <sstream> #include <fstream> #if !defined(CUSTOM_STACKTRACE_IMPL) && BOOST_VERSION >= 106500 #include <boost/stacktrace.hpp> namespace server_lib { void emergency_helper::save_dump(const char* dump_file_path) { try { boost::stacktrace::safe_dump_to(dump_file_path); } catch (const std::exception& e) { std::cerr << "Can't save dump: " << e.what() << std::endl; } } std::string emergency_helper::load_dump(const char* dump_file_path, bool remove) { std::stringstream ss; if (boost::filesystem::exists(dump_file_path)) { try { std::ifstream ifs(dump_file_path); boost::stacktrace::stacktrace st = boost::stacktrace::stacktrace::from_dump(ifs); ss << st; // cleaning up ifs.close(); } catch (const std::exception& e) { ss << e.what(); } ss << '\n'; try { if (remove) boost::filesystem::remove(dump_file_path); } catch (const std::exception&) { // ignore } } return ss.str(); } } // namespace server_lib #else #include <stdlib.h> #include <execinfo.h> #include <cxxabi.h> namespace server_lib { void emergency_helper::save_dump(const char* dump_file_path) { FILE* out = fopen(dump_file_path, "w"); if (!out) { perror("Can't save dump"); return; } /** Print a demangled stack backtrace of the caller function to FILE* out. */ const size_t max_frames = 100; fprintf(out, "stack trace:\n"); // storage array for stack trace address data void* addrlist[max_frames + 1] = {}; // retrieve current stack addresses: // https://linux.die.net/man/3/backtrace_symbols: //"The symbol names may be unavailable without the use of special linker options. // For systems using the GNU linker, it is necessary to use the -rdynamic linker option. // Note that names of "static" functions are not exposed, and won't be available in the backtrace." // int addrlen = backtrace(addrlist, sizeof(addrlist) / sizeof(void*)); if (addrlen == 0) { fprintf(out, " <empty, possibly corrupt>\n"); return; } // resolve addresses into strings containing "filename(function+address)", // this array must be free()-ed char** symbollist = backtrace_symbols(addrlist, addrlen); // allocate string which will be filled with the demangled function name size_t funcnamesize = 256; char* funcname = reinterpret_cast<char*>(alloca(funcnamesize)); // iterate over the returned symbol lines. skip the first, it is the // address of this function. for (int i = 1; i < addrlen; i++) { char *begin_name = nullptr, *begin_offset = nullptr, *end_offset = nullptr; // find parentheses and +address offset surrounding the mangled name: // ./module(function+0x15c) [0x8048a6d] for (char* p = symbollist[i]; *p; ++p) { if (*p == '(') begin_name = p; else if (*p == '+') begin_offset = p; else if (*p == ')' && begin_offset) { end_offset = p; break; } } if (begin_name && begin_offset && end_offset && begin_name < begin_offset) { *begin_name++ = '\0'; *begin_offset++ = '\0'; *end_offset = '\0'; // mangled name is now in [begin_name, begin_offset) and caller // offset in [begin_offset, end_offset). now apply // __cxa_demangle(): int status; char* ret = abi::__cxa_demangle(begin_name, funcname, &funcnamesize, &status); if (status == 0) { funcname = ret; // use possibly realloc()-ed string fprintf(out, " %s : %s+%s\n", symbollist[i], funcname, begin_offset); } else { // demangling failed. Output function name as a C function with // no arguments. fprintf(out, " %s : %s()+%s\n", symbollist[i], begin_name, begin_offset); } } else { // couldn't parse the line? print the whole line. fprintf(out, " %s\n", symbollist[i]); } } fclose(out); free(symbollist); } std::string emergency_helper::load_dump(const char* dump_file_path, bool remove) { if (dump_file_path && boost::filesystem::exists(dump_file_path)) { std::stringstream ss; try { std::ifstream ifs(dump_file_path); std::string line; while (std::getline(ifs, line)) { ss << line << '\n'; } // cleaning up ifs.close(); } catch (const std::exception& e) { ss << e.what(); } ss << '\n'; try { if (remove) boost::filesystem::remove(dump_file_path); } catch (const std::exception&) { // ignore } return ss.str(); } else return {}; } } // namespace server_lib #endif namespace server_lib { bool emergency_helper::test_for_write(const char* dump_file_path) { if (boost::filesystem::exists(dump_file_path)) return false; { std::ofstream ofs(dump_file_path); if (!ofs) return false; } boost::filesystem::remove(dump_file_path); return true; } } // namespace server_lib #else //!STACKTRACE_DISABLED // raise linker error for save_dump, load_dump #endif
{ "alphanum_fraction": 0.5386509109, "author": null, "avg_line_length": 25.3875, "converted": null, "ext": "cpp", "file": null, "hexsha": "33b042f69c5293fe11643ebbed12756cfdac17dd", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-08-16T13:38:15.000Z", "max_forks_repo_forks_event_min_datetime": "2021-08-16T13:38:15.000Z", "max_forks_repo_head_hexsha": "d0778f08958b6f2d37f67d577c98ebf46c42b2b9", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "romualdo-bar/barbacoa-server-lib", "max_forks_repo_path": "src/emergency_helper.cpp", "max_issues_count": null, "max_issues_repo_head_hexsha": "d0778f08958b6f2d37f67d577c98ebf46c42b2b9", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "romualdo-bar/barbacoa-server-lib", "max_issues_repo_path": "src/emergency_helper.cpp", "max_line_length": 103, "max_stars_count": 1, "max_stars_repo_head_hexsha": "d0778f08958b6f2d37f67d577c98ebf46c42b2b9", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "romualdo-bar/barbacoa-server-lib", "max_stars_repo_path": "src/emergency_helper.cpp", "max_stars_repo_stars_event_max_datetime": "2020-10-27T19:23:17.000Z", "max_stars_repo_stars_event_min_datetime": "2020-10-27T19:23:17.000Z", "num_tokens": 1394, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 6093 }
""" Algorithms that Involve Multiple DataFrames =========================================== The pandas operations ``concat``, ``join``, and ``merge`` combine multiple DataFrames. This module contains analogous algorithms in the parallel case. There are two important cases: 1. We combine along a partitioned index 2. We combine along an unpartitioned index or other column In the first case we know which partitions of each dataframe interact with which others. This lets us be significantly more clever and efficient. In the second case each partition from one dataset interacts with all partitions from the other. We handle this through a shuffle operation. Partitioned Joins ----------------- In the first case where we join along a partitioned index we proceed in the following stages. 1. Align the partitions of all inputs to be the same. This involves a call to ``dd.repartition`` which will split up and concat existing partitions as necessary. After this step all inputs have partitions that align with each other. This step is relatively cheap. See the function ``align_partitions``. 2. Remove unnecessary partitions based on the type of join we perform (left, right, inner, outer). We can do this at the partition level before any computation happens. We'll do it again on each partition when we call the in-memory function. See the function ``require``. 3. Embarrassingly parallel calls to ``pd.concat``, ``pd.join``, or ``pd.merge``. Now that the data is aligned and unnecessary blocks have been removed we can rely on the fast in-memory Pandas join machinery to execute joins per-partition. We know that all intersecting records exist within the same partition Hash Joins via Shuffle ---------------------- When we join along an unpartitioned index or along an arbitrary column any partition from one input might interact with any partition in another. In this case we perform a hash-join by shuffling data in each input by that column. This results in new inputs with the same partition structure cleanly separated along that column. We proceed with hash joins in the following stages: 1. Shuffle each input on the specified column. See the function ``dask.dataframe.shuffle.shuffle``. 2. Perform embarrassingly parallel join across shuffled inputs. """ import math import pickle import warnings from functools import partial, wraps import numpy as np import pandas as pd from pandas.api.types import is_categorical_dtype, is_dtype_equal from tlz import merge_sorted, unique from dask.base import is_dask_collection, tokenize from dask.dataframe import methods from dask.dataframe.core import ( DataFrame, Index, Series, _concat, _Frame, _maybe_from_pandas, is_broadcastable, map_partitions, new_dd_object, prefix_reduction, suffix_reduction, ) from dask.dataframe.dispatch import group_split_dispatch, hash_object_dispatch from dask.dataframe.io import from_pandas from dask.dataframe.shuffle import ( partitioning_index, rearrange_by_divisions, shuffle, shuffle_group, ) from dask.dataframe.utils import ( asciitable, is_dataframe_like, is_series_like, make_meta, strip_unknown_categories, ) from dask.highlevelgraph import HighLevelGraph from dask.layers import BroadcastJoinLayer from dask.utils import M, apply def align_partitions(*dfs): """Mutually partition and align DataFrame blocks This serves as precursor to multi-dataframe operations like join, concat, or merge. Parameters ---------- dfs: sequence of dd.DataFrame, dd.Series and dd.base.Scalar Sequence of dataframes to be aligned on their index Returns ------- dfs: sequence of dd.DataFrame, dd.Series and dd.base.Scalar These must have consistent divisions with each other divisions: tuple Full divisions sequence of the entire result result: list A list of lists of keys that show which data exist on which divisions """ _is_broadcastable = partial(is_broadcastable, dfs) dfs1 = [df for df in dfs if isinstance(df, _Frame) and not _is_broadcastable(df)] if len(dfs) == 0: raise ValueError("dfs contains no DataFrame and Series") if not all(df.known_divisions for df in dfs1): raise ValueError( "Not all divisions are known, can't align " "partitions. Please use `set_index` " "to set the index." ) divisions = list(unique(merge_sorted(*[df.divisions for df in dfs1]))) if len(divisions) == 1: # single value for index divisions = (divisions[0], divisions[0]) dfs2 = [ df.repartition(divisions, force=True) if isinstance(df, _Frame) else df for df in dfs ] result = list() inds = [0 for df in dfs] for d in divisions[:-1]: L = list() for i, df in enumerate(dfs2): if isinstance(df, _Frame): j = inds[i] divs = df.divisions if j < len(divs) - 1 and divs[j] == d: L.append((df._name, inds[i])) inds[i] += 1 else: L.append(None) else: # Scalar has no divisions L.append(None) result.append(L) return dfs2, tuple(divisions), result def _maybe_align_partitions(args): """Align DataFrame blocks if divisions are different. Note that if all divisions are unknown, but have equal npartitions, then they will be passed through unchanged. This is different than `align_partitions`, which will fail if divisions aren't all known""" _is_broadcastable = partial(is_broadcastable, args) dfs = [df for df in args if isinstance(df, _Frame) and not _is_broadcastable(df)] if not dfs: return args divisions = dfs[0].divisions if not all(df.divisions == divisions for df in dfs): dfs2 = iter(align_partitions(*dfs)[0]) return [a if not isinstance(a, _Frame) else next(dfs2) for a in args] return args def require(divisions, parts, required=None): """Clear out divisions where required components are not present In left, right, or inner joins we exclude portions of the dataset if one side or the other is not present. We can achieve this at the partition level as well >>> divisions = [1, 3, 5, 7, 9] >>> parts = [(('a', 0), None), ... (('a', 1), ('b', 0)), ... (('a', 2), ('b', 1)), ... (None, ('b', 2))] >>> divisions2, parts2 = require(divisions, parts, required=[0]) >>> divisions2 (1, 3, 5, 7) >>> parts2 # doctest: +NORMALIZE_WHITESPACE ((('a', 0), None), (('a', 1), ('b', 0)), (('a', 2), ('b', 1))) >>> divisions2, parts2 = require(divisions, parts, required=[1]) >>> divisions2 (3, 5, 7, 9) >>> parts2 # doctest: +NORMALIZE_WHITESPACE ((('a', 1), ('b', 0)), (('a', 2), ('b', 1)), (None, ('b', 2))) >>> divisions2, parts2 = require(divisions, parts, required=[0, 1]) >>> divisions2 (3, 5, 7) >>> parts2 # doctest: +NORMALIZE_WHITESPACE ((('a', 1), ('b', 0)), (('a', 2), ('b', 1))) """ if not required: return divisions, parts for i in required: present = [j for j, p in enumerate(parts) if p[i] is not None] divisions = tuple(divisions[min(present) : max(present) + 2]) parts = tuple(parts[min(present) : max(present) + 1]) return divisions, parts ############################################################### # Join / Merge ############################################################### required = { "left": [0], "leftsemi": [0], "leftanti": [0], "right": [1], "inner": [0, 1], "outer": [], } allowed_left = ("inner", "left", "leftsemi", "leftanti") allowed_right = ("inner", "right") def merge_chunk(lhs, *args, empty_index_dtype=None, categorical_columns=None, **kwargs): rhs, *args = args left_index = kwargs.get("left_index", False) right_index = kwargs.get("right_index", False) if categorical_columns is not None: for col in categorical_columns: left = None right = None if col in lhs: left = lhs[col] elif col == kwargs.get("right_on", None) and left_index: if is_categorical_dtype(lhs.index): left = lhs.index if col in rhs: right = rhs[col] elif col == kwargs.get("left_on", None) and right_index: if is_categorical_dtype(rhs.index): right = rhs.index dtype = "category" if left is not None and right is not None: dtype = methods.union_categoricals( [left.astype("category"), right.astype("category")] ).dtype if left is not None: if isinstance(left, pd.Index): lhs.index = left.astype(dtype) else: lhs = lhs.assign(**{col: left.astype(dtype)}) if right is not None: if isinstance(right, pd.Index): rhs.index = right.astype(dtype) else: rhs = rhs.assign(**{col: right.astype(dtype)}) out = lhs.merge(rhs, *args, **kwargs) # Workaround pandas bug where if the output result of a merge operation is # an empty dataframe, the output index is `int64` in all cases, regardless # of input dtypes. if len(out) == 0 and empty_index_dtype is not None: out.index = out.index.astype(empty_index_dtype) return out def merge_indexed_dataframes(lhs, rhs, left_index=True, right_index=True, **kwargs): """Join two partitioned dataframes along their index""" how = kwargs.get("how", "left") kwargs["left_index"] = left_index kwargs["right_index"] = right_index (lhs, rhs), divisions, parts = align_partitions(lhs, rhs) divisions, parts = require(divisions, parts, required[how]) name = "join-indexed-" + tokenize(lhs, rhs, **kwargs) meta = lhs._meta_nonempty.merge(rhs._meta_nonempty, **kwargs) kwargs["empty_index_dtype"] = meta.index.dtype kwargs["categorical_columns"] = meta.select_dtypes(include="category").columns dsk = dict() for i, (a, b) in enumerate(parts): dsk[(name, i)] = (apply, merge_chunk, [a, b], kwargs) graph = HighLevelGraph.from_collections(name, dsk, dependencies=[lhs, rhs]) return new_dd_object(graph, name, meta, divisions) shuffle_func = shuffle # name sometimes conflicts with keyword argument def hash_join( lhs, left_on, rhs, right_on, how="inner", npartitions=None, suffixes=("_x", "_y"), shuffle=None, indicator=False, max_branch=None, ): """Join two DataFrames on particular columns with hash join This shuffles both datasets on the joined column and then performs an embarrassingly parallel join partition-by-partition >>> hash_join(lhs, 'id', rhs, 'id', how='left', npartitions=10) # doctest: +SKIP """ if npartitions is None: npartitions = max(lhs.npartitions, rhs.npartitions) lhs2 = shuffle_func( lhs, left_on, npartitions=npartitions, shuffle=shuffle, max_branch=max_branch ) rhs2 = shuffle_func( rhs, right_on, npartitions=npartitions, shuffle=shuffle, max_branch=max_branch ) if isinstance(left_on, Index): left_on = None left_index = True else: left_index = False if isinstance(right_on, Index): right_on = None right_index = True else: right_index = False kwargs = dict( how=how, left_on=left_on, right_on=right_on, left_index=left_index, right_index=right_index, suffixes=suffixes, indicator=indicator, ) # dummy result # Avoid using dummy data for a collection it is empty _lhs_meta = lhs._meta_nonempty if len(lhs.columns) else lhs._meta _rhs_meta = rhs._meta_nonempty if len(rhs.columns) else rhs._meta meta = _lhs_meta.merge(_rhs_meta, **kwargs) if isinstance(left_on, list): left_on = (list, tuple(left_on)) if isinstance(right_on, list): right_on = (list, tuple(right_on)) kwargs["empty_index_dtype"] = meta.index.dtype kwargs["categorical_columns"] = meta.select_dtypes(include="category").columns joined = map_partitions( merge_chunk, lhs2, rhs2, meta=meta, enforce_metadata=False, transform_divisions=False, align_dataframes=False, **kwargs, ) return joined def single_partition_join(left, right, **kwargs): # if the merge is performed on_index, divisions can be kept, otherwise the # new index will not necessarily correspond with the current divisions meta = left._meta_nonempty.merge(right._meta_nonempty, **kwargs) use_left = kwargs.get("right_index") or right._contains_index_name( kwargs.get("right_on") ) use_right = kwargs.get("left_index") or left._contains_index_name( kwargs.get("left_on") ) if len(meta) == 0: if use_left: meta.index = meta.index.astype(left.index.dtype) elif use_right: meta.index = meta.index.astype(right.index.dtype) else: meta.index = meta.index.astype("int64") kwargs["empty_index_dtype"] = meta.index.dtype kwargs["categorical_columns"] = meta.select_dtypes(include="category").columns if right.npartitions == 1 and kwargs["how"] in allowed_left: if use_left: divisions = left.divisions elif use_right and len(right.divisions) == len(left.divisions): divisions = right.divisions else: divisions = [None for _ in left.divisions] elif left.npartitions == 1 and kwargs["how"] in allowed_right: if use_right: divisions = right.divisions elif use_left and len(left.divisions) == len(right.divisions): divisions = left.divisions else: divisions = [None for _ in right.divisions] else: raise NotImplementedError( "single_partition_join has no fallback for invalid calls" ) joined = map_partitions( merge_chunk, left, right, meta=meta, enforce_metadata=False, transform_divisions=False, align_dataframes=False, **kwargs, ) joined.divisions = tuple(divisions) return joined def warn_dtype_mismatch(left, right, left_on, right_on): """Checks for merge column dtype mismatches and throws a warning (#4574)""" if not isinstance(left_on, list): left_on = [left_on] if not isinstance(right_on, list): right_on = [right_on] if all(col in left.columns for col in left_on) and all( col in right.columns for col in right_on ): dtype_mism = [ ((lo, ro), left.dtypes[lo], right.dtypes[ro]) for lo, ro in zip(left_on, right_on) if not is_dtype_equal(left.dtypes[lo], right.dtypes[ro]) ] if dtype_mism: col_tb = asciitable( ("Merge columns", "left dtype", "right dtype"), dtype_mism ) warnings.warn( ( "Merging dataframes with merge column data " "type mismatches: \n{}\nCast dtypes explicitly to " "avoid unexpected results." ).format(col_tb) ) @wraps(pd.merge) def merge( left, right, how="inner", on=None, left_on=None, right_on=None, left_index=False, right_index=False, suffixes=("_x", "_y"), indicator=False, npartitions=None, shuffle=None, max_branch=None, broadcast=None, ): for o in [on, left_on, right_on]: if isinstance(o, _Frame): raise NotImplementedError( "Dask collections not currently allowed in merge columns" ) if not on and not left_on and not right_on and not left_index and not right_index: on = [c for c in left.columns if c in right.columns] if not on: left_index = right_index = True if on and not left_on and not right_on: left_on = right_on = on on = None supported_how = ("left", "right", "outer", "inner", "leftanti", "leftsemi") if how not in supported_how: raise ValueError( f"dask.dataframe.merge does not support how='{how}'. Options are: {supported_how}." f" Note that 'leftanti' and 'leftsemi' are only dask_cudf options." ) if isinstance(left, (pd.Series, pd.DataFrame)) and isinstance( right, (pd.Series, pd.DataFrame) ): return pd.merge( left, right, how=how, on=on, left_on=left_on, right_on=right_on, left_index=left_index, right_index=right_index, suffixes=suffixes, indicator=indicator, ) # Transform pandas objects into dask.dataframe objects if not is_dask_collection(left): if right_index and left_on: # change to join on index left = left.set_index(left[left_on]) left_on = None left_index = True left = from_pandas(left, npartitions=1) # turn into DataFrame if not is_dask_collection(right): if left_index and right_on: # change to join on index right = right.set_index(right[right_on]) right_on = None right_index = True right = from_pandas(right, npartitions=1) # turn into DataFrame # Both sides are now dd.DataFrame or dd.Series objects merge_indexed_left = ( left_index or left._contains_index_name(left_on) ) and left.known_divisions merge_indexed_right = ( right_index or right._contains_index_name(right_on) ) and right.known_divisions # Both sides indexed if merge_indexed_left and merge_indexed_right: # Do indexed join return merge_indexed_dataframes( left, right, how=how, suffixes=suffixes, indicator=indicator, left_on=left_on, right_on=right_on, left_index=left_index, right_index=right_index, ) # Single partition on one side # Note that cudf supports "leftsemi" and "leftanti" joins elif ( left.npartitions == 1 and how in allowed_right or right.npartitions == 1 and how in allowed_left ): return single_partition_join( left, right, how=how, right_on=right_on, left_on=left_on, left_index=left_index, right_index=right_index, suffixes=suffixes, indicator=indicator, ) # One side is indexed, the other not elif ( left_index and left.known_divisions and not right_index or right_index and right.known_divisions and not left_index ): left_empty = left._meta_nonempty right_empty = right._meta_nonempty meta = left_empty.merge( right_empty, how=how, on=on, left_on=left_on, right_on=right_on, left_index=left_index, right_index=right_index, suffixes=suffixes, indicator=indicator, ) categorical_columns = meta.select_dtypes(include="category").columns if merge_indexed_left and left.known_divisions: right = rearrange_by_divisions( right, right_on, left.divisions, max_branch, shuffle=shuffle ) left = left.clear_divisions() elif merge_indexed_right and right.known_divisions: left = rearrange_by_divisions( left, left_on, right.divisions, max_branch, shuffle=shuffle ) right = right.clear_divisions() return map_partitions( merge_chunk, left, right, meta=meta, how=how, on=on, left_on=left_on, right_on=right_on, left_index=left_index, right_index=right_index, suffixes=suffixes, indicator=indicator, empty_index_dtype=meta.index.dtype, categorical_columns=categorical_columns, ) # Catch all hash join else: if left_on and right_on: warn_dtype_mismatch(left, right, left_on, right_on) # Check if we should use a broadcast_join # See note on `broadcast_bias` below. broadcast_bias = 0.5 if isinstance(broadcast, float): broadcast_bias = float(broadcast) broadcast = None elif not isinstance(broadcast, bool) and broadcast is not None: # Let's be strict about the `broadcast` type to # avoid arbitrarily casting int to float or bool. raise ValueError( f"Optional `broadcast` argument must be float or bool." f"Type={type(broadcast)} is not supported." ) bcast_side = "left" if left.npartitions < right.npartitions else "right" n_small = min(left.npartitions, right.npartitions) n_big = max(left.npartitions, right.npartitions) if ( shuffle == "tasks" and how in ("inner", "left", "right") and how != bcast_side and broadcast is not False ): # Note on `broadcast_bias`: # We can expect the broadcast merge to be competitive with # the shuffle merge when the number of partitions in the # smaller collection is less than the logarithm of the number # of partitions in the larger collection. By default, we add # a small preference for the shuffle-based merge by multiplying # the log result by a 0.5 scaling factor. We call this factor # the `broadcast_bias`, because a larger number will make Dask # more likely to select the `broadcast_join` code path. If # the user specifies a floating-point value for the `broadcast` # kwarg, that value will be used as the `broadcast_bias`. if broadcast or (n_small < math.log2(n_big) * broadcast_bias): return broadcast_join( left, left.index if left_index else left_on, right, right.index if right_index else right_on, how, npartitions, suffixes, indicator=indicator, ) return hash_join( left, left.index if left_index else left_on, right, right.index if right_index else right_on, how, npartitions, suffixes, shuffle=shuffle, indicator=indicator, max_branch=max_branch, ) ############################################################### # ASOF Join ############################################################### def most_recent_tail(left, right): if len(right.index) == 0: return left return right.tail(1) def most_recent_tail_summary(left, right, by=None): return pd.concat([left, right]).drop_duplicates(subset=by, keep="last") def compute_tails(ddf, by=None): """For each partition, returns the last row of the most recent nonempty partition. """ empty = ddf._meta.iloc[0:0] if by is None: return prefix_reduction(most_recent_tail, ddf, empty) else: kwargs = {"by": by} return prefix_reduction(most_recent_tail_summary, ddf, empty, **kwargs) def most_recent_head(left, right): if len(left.index) == 0: return right return left.head(1) def most_recent_head_summary(left, right, by=None): return pd.concat([left, right]).drop_duplicates(subset=by, keep="first") def compute_heads(ddf, by=None): """For each partition, returns the first row of the next nonempty partition. """ empty = ddf._meta.iloc[0:0] if by is None: return suffix_reduction(most_recent_head, ddf, empty) else: kwargs = {"by": by} return suffix_reduction(most_recent_head_summary, ddf, empty, **kwargs) def pair_partitions(L, R): """Returns which partitions to pair for the merge_asof algorithm and the bounds on which to split them up """ result = [] n, m = len(L) - 1, len(R) - 1 i, j = 0, -1 while j + 1 < m and R[j + 1] <= L[i]: j += 1 J = [] while i < n: partition = max(0, min(m - 1, j)) lower = R[j] if j >= 0 and R[j] > L[i] else None upper = ( R[j + 1] if j + 1 < m and (R[j + 1] < L[i + 1] or R[j + 1] == L[i + 1] and i == n - 1) else None ) J.append((partition, lower, upper)) i1 = i + 1 if j + 1 == m or (i + 1 < n and R[j + 1] >= L[i + 1]) else i j1 = j + 1 if i + 1 == n or (j + 1 < m and L[i + 1] >= R[j + 1]) else j if i1 > i: result.append(J) J = [] elif i == n - 1 and R[j1] > L[n]: result.append(J) break i, j = i1, j1 return result def merge_asof_padded(left, right, prev=None, next=None, **kwargs): """merge_asof but potentially adding rows to the beginning/end of right""" frames = [] if prev is not None: frames.append(prev) frames.append(right) if next is not None: frames.append(next) frame = pd.concat(frames) result = pd.merge_asof(left, frame, **kwargs) # pd.merge_asof() resets index name (and dtype) if left is empty df if result.index.name != left.index.name: result.index.name = left.index.name return result def get_unsorted_columns(frames): """ Determine the unsorted column order. This should match the output of concat([frames], sort=False) """ new_columns = pd.concat([frame._meta for frame in frames]).columns order = [] for frame in frames: order.append(new_columns.get_indexer_for(frame.columns)) order = np.concatenate(order) order = pd.unique(order) order = new_columns.take(order) return order def merge_asof_indexed(left, right, **kwargs): dsk = dict() name = "asof-join-indexed-" + tokenize(left, right, **kwargs) meta = pd.merge_asof(left._meta_nonempty, right._meta_nonempty, **kwargs) if all(map(pd.isnull, left.divisions)): # results in an empty df that looks like ``meta`` return from_pandas(meta.iloc[len(meta) :], npartitions=left.npartitions) if all(map(pd.isnull, right.divisions)): # results in an df that looks like ``left`` with nulls for # all ``right.columns`` return map_partitions( pd.merge_asof, left, right=right, left_index=True, right_index=True, meta=meta, ) dependencies = [left, right] tails = heads = None if kwargs["direction"] in ["backward", "nearest"]: tails = compute_tails(right, by=kwargs["right_by"]) dependencies.append(tails) if kwargs["direction"] in ["forward", "nearest"]: heads = compute_heads(right, by=kwargs["right_by"]) dependencies.append(heads) for i, J in enumerate(pair_partitions(left.divisions, right.divisions)): frames = [] for j, lower, upper in J: slice = (methods.boundary_slice, (left._name, i), lower, upper, False) tail = (tails._name, j) if tails is not None else None head = (heads._name, j) if heads is not None else None frames.append( ( apply, merge_asof_padded, [slice, (right._name, j), tail, head], kwargs, ) ) dsk[(name, i)] = (methods.concat, frames) graph = HighLevelGraph.from_collections(name, dsk, dependencies=dependencies) result = new_dd_object(graph, name, meta, left.divisions) return result @wraps(pd.merge_asof) def merge_asof( left, right, on=None, left_on=None, right_on=None, left_index=False, right_index=False, by=None, left_by=None, right_by=None, suffixes=("_x", "_y"), tolerance=None, allow_exact_matches=True, direction="backward", ): if direction not in ["backward", "forward", "nearest"]: raise ValueError( "Invalid merge_asof direction. Choose from 'backward'" " 'forward', or 'nearest'" ) kwargs = { "on": on, "left_on": left_on, "right_on": right_on, "left_index": left_index, "right_index": right_index, "by": by, "left_by": left_by, "right_by": right_by, "suffixes": suffixes, "tolerance": tolerance, "allow_exact_matches": allow_exact_matches, "direction": direction, } if left is None or right is None: raise ValueError("Cannot merge_asof on None") # if is_dataframe_like(left) and is_dataframe_like(right): if isinstance(left, pd.DataFrame) and isinstance(right, pd.DataFrame): return pd.merge_asof(left, right, **kwargs) if on is not None: if left_on is not None or right_on is not None: raise ValueError( "Can only pass argument 'on' OR 'left_on' and 'right_on', not a " "combination of both." ) left_on = right_on = on for o in [left_on, right_on]: if isinstance(o, _Frame): raise NotImplementedError( "Dask collections not currently allowed in merge columns" ) if not is_dask_collection(left): left = from_pandas(left, npartitions=1) ixname = ixcol = divs = None if left_on is not None: if right_index: divs = left.divisions if left.known_divisions else None ixname = left.index.name left = left.reset_index() ixcol = left.columns[0] left = left.set_index(left_on, sorted=True) if not is_dask_collection(right): right = from_pandas(right, npartitions=1) if right_on is not None: right = right.set_index(right_on, drop=(left_on == right_on), sorted=True) if by is not None: if left_by is not None or right_by is not None: raise ValueError( "Can only pass argument 'by' OR 'left_by' and 'right_by', not a combination of both." ) kwargs["left_by"] = kwargs["right_by"] = by if left_by is None and right_by is not None: raise ValueError("Must specify both left_on and right_on if one is specified.") if left_by is not None and right_by is None: raise ValueError("Must specify both left_on and right_on if one is specified.") del kwargs["on"], kwargs["left_on"], kwargs["right_on"], kwargs["by"] kwargs["left_index"] = kwargs["right_index"] = True if not left.known_divisions or not right.known_divisions: raise ValueError("merge_asof input must be sorted!") result = merge_asof_indexed(left, right, **kwargs) if left_on or right_on: result = result.reset_index() if ixcol is not None: if divs is not None: result = result.set_index(ixcol, sorted=True, divisions=divs) else: result = result.map_partitions(M.set_index, ixcol) result = result.map_partitions(M.rename_axis, ixname) return result ############################################################### # Concat ############################################################### def concat_and_check(dfs, ignore_order=False): if len(set(map(len, dfs))) != 1: raise ValueError("Concatenated DataFrames of different lengths") return methods.concat(dfs, axis=1, ignore_order=ignore_order) def concat_unindexed_dataframes(dfs, ignore_order=False, **kwargs): name = "concat-" + tokenize(*dfs) dsk = { (name, i): (concat_and_check, [(df._name, i) for df in dfs], ignore_order) for i in range(dfs[0].npartitions) } kwargs.update({"ignore_order": ignore_order}) meta = methods.concat([df._meta for df in dfs], axis=1, **kwargs) graph = HighLevelGraph.from_collections(name, dsk, dependencies=dfs) return new_dd_object(graph, name, meta, dfs[0].divisions) def concat_indexed_dataframes(dfs, axis=0, join="outer", ignore_order=False, **kwargs): """Concatenate indexed dataframes together along the index""" warn = axis != 0 kwargs.update({"ignore_order": ignore_order}) meta = methods.concat( [df._meta for df in dfs], axis=axis, join=join, filter_warning=warn, **kwargs, ) empties = [strip_unknown_categories(df._meta) for df in dfs] dfs2, divisions, parts = align_partitions(*dfs) name = "concat-indexed-" + tokenize(join, *dfs) parts2 = [ [df if df is not None else empty for df, empty in zip(part, empties)] for part in parts ] filter_warning = True uniform = False dsk = { (name, i): (methods.concat, part, axis, join, uniform, filter_warning, kwargs) for i, part in enumerate(parts2) } for df in dfs2: dsk.update(df.dask) return new_dd_object(dsk, name, meta, divisions) def stack_partitions(dfs, divisions, join="outer", ignore_order=False, **kwargs): """Concatenate partitions on axis=0 by doing a simple stack""" # Use _meta_nonempty as pandas.concat will incorrectly cast float to datetime # for empty data frames. See https://github.com/pandas-dev/pandas/issues/32934. kwargs.update({"ignore_order": ignore_order}) meta = make_meta( methods.concat( [df._meta_nonempty for df in dfs], join=join, filter_warning=False, **kwargs, ) ) empty = strip_unknown_categories(meta) name = f"concat-{tokenize(*dfs)}" dsk = {} i = 0 astyped_dfs = [] for df in dfs: # dtypes of all dfs need to be coherent # refer to https://github.com/dask/dask/issues/4685 # and https://github.com/dask/dask/issues/5968. if is_dataframe_like(df): shared_columns = df.columns.intersection(meta.columns) needs_astype = [ col for col in shared_columns if df[col].dtype != meta[col].dtype and not is_categorical_dtype(df[col].dtype) ] if needs_astype: # Copy to avoid mutating the caller inplace df = df.copy() df[needs_astype] = df[needs_astype].astype(meta[needs_astype].dtypes) if is_series_like(df) and is_series_like(meta): if not df.dtype == meta.dtype and not is_categorical_dtype(df.dtype): df = df.astype(meta.dtype) else: pass # TODO: there are other non-covered cases here astyped_dfs.append(df) # An error will be raised if the schemas or categories don't match. In # this case we need to pass along the meta object to transform each # partition, so they're all equivalent. try: df._meta == meta match = True except (ValueError, TypeError): match = False filter_warning = True uniform = False for key in df.__dask_keys__(): if match: dsk[(name, i)] = key else: dsk[(name, i)] = ( apply, methods.concat, [[empty, key], 0, join, uniform, filter_warning], kwargs, ) i += 1 graph = HighLevelGraph.from_collections(name, dsk, dependencies=astyped_dfs) return new_dd_object(graph, name, meta, divisions) def concat( dfs, axis=0, join="outer", interleave_partitions=False, ignore_unknown_divisions=False, ignore_order=False, **kwargs, ): """Concatenate DataFrames along rows. - When axis=0 (default), concatenate DataFrames row-wise: - If all divisions are known and ordered, concatenate DataFrames keeping divisions. When divisions are not ordered, specifying interleave_partition=True allows concatenate divisions each by each. - If any of division is unknown, concatenate DataFrames resetting its division to unknown (None) - When axis=1, concatenate DataFrames column-wise: - Allowed if all divisions are known. - If any of division is unknown, it raises ValueError. Parameters ---------- dfs : list List of dask.DataFrames to be concatenated axis : {0, 1, 'index', 'columns'}, default 0 The axis to concatenate along join : {'inner', 'outer'}, default 'outer' How to handle indexes on other axis interleave_partitions : bool, default False Whether to concatenate DataFrames ignoring its order. If True, every divisions are concatenated each by each. ignore_unknown_divisions : bool, default False By default a warning is raised if any input has unknown divisions. Set to True to disable this warning. ignore_order : bool, default False Whether to ignore order when doing the union of categoricals. Notes ----- This differs in from ``pd.concat`` in the when concatenating Categoricals with different categories. Pandas currently coerces those to objects before concatenating. Coercing to objects is very expensive for large arrays, so dask preserves the Categoricals by taking the union of the categories. Examples -------- If all divisions are known and ordered, divisions are kept. >>> import dask.dataframe as dd >>> a # doctest: +SKIP dd.DataFrame<x, divisions=(1, 3, 5)> >>> b # doctest: +SKIP dd.DataFrame<y, divisions=(6, 8, 10)> >>> dd.concat([a, b]) # doctest: +SKIP dd.DataFrame<concat-..., divisions=(1, 3, 6, 8, 10)> Unable to concatenate if divisions are not ordered. >>> a # doctest: +SKIP dd.DataFrame<x, divisions=(1, 3, 5)> >>> b # doctest: +SKIP dd.DataFrame<y, divisions=(2, 3, 6)> >>> dd.concat([a, b]) # doctest: +SKIP ValueError: All inputs have known divisions which cannot be concatenated in order. Specify interleave_partitions=True to ignore order Specify interleave_partitions=True to ignore the division order. >>> dd.concat([a, b], interleave_partitions=True) # doctest: +SKIP dd.DataFrame<concat-..., divisions=(1, 2, 3, 5, 6)> If any of division is unknown, the result division will be unknown >>> a # doctest: +SKIP dd.DataFrame<x, divisions=(None, None)> >>> b # doctest: +SKIP dd.DataFrame<y, divisions=(1, 4, 10)> >>> dd.concat([a, b]) # doctest: +SKIP dd.DataFrame<concat-..., divisions=(None, None, None, None)> By default concatenating with unknown divisions will raise a warning. Set ``ignore_unknown_divisions=True`` to disable this: >>> dd.concat([a, b], ignore_unknown_divisions=True)# doctest: +SKIP dd.DataFrame<concat-..., divisions=(None, None, None, None)> Different categoricals are unioned >>> dd.concat([ ... dd.from_pandas(pd.Series(['a', 'b'], dtype='category'), 1), ... dd.from_pandas(pd.Series(['a', 'c'], dtype='category'), 1), ... ], interleave_partitions=True).dtype CategoricalDtype(categories=['a', 'b', 'c'], ordered=False) """ if not isinstance(dfs, list): raise TypeError("dfs must be a list of DataFrames/Series objects") if len(dfs) == 0: raise ValueError("No objects to concatenate") if len(dfs) == 1: if axis == 1 and isinstance(dfs[0], Series): return dfs[0].to_frame() else: return dfs[0] if join not in ("inner", "outer"): raise ValueError("'join' must be 'inner' or 'outer'") axis = DataFrame._validate_axis(axis) dasks = [df for df in dfs if isinstance(df, _Frame)] dfs = _maybe_from_pandas(dfs) if axis == 1: if all(df.known_divisions for df in dasks): return concat_indexed_dataframes( dfs, axis=axis, join=join, ignore_order=ignore_order, **kwargs ) elif ( len(dasks) == len(dfs) and all(not df.known_divisions for df in dfs) and len({df.npartitions for df in dasks}) == 1 ): if not ignore_unknown_divisions: warnings.warn( "Concatenating dataframes with unknown divisions.\n" "We're assuming that the indices of each dataframes" " are \n aligned. This assumption is not generally " "safe." ) return concat_unindexed_dataframes(dfs, ignore_order=ignore_order, **kwargs) else: raise ValueError( "Unable to concatenate DataFrame with unknown " "division specifying axis=1" ) else: if all(df.known_divisions for df in dasks): # each DataFrame's division must be greater than previous one if all( dfs[i].divisions[-1] < dfs[i + 1].divisions[0] for i in range(len(dfs) - 1) ): divisions = [] for df in dfs[:-1]: # remove last to concatenate with next divisions += df.divisions[:-1] divisions += dfs[-1].divisions return stack_partitions( dfs, divisions, join=join, ignore_order=ignore_order, **kwargs ) elif interleave_partitions: return concat_indexed_dataframes( dfs, join=join, ignore_order=ignore_order, **kwargs ) else: divisions = [None] * (sum(df.npartitions for df in dfs) + 1) return stack_partitions( dfs, divisions, join=join, ignore_order=ignore_order, **kwargs ) else: divisions = [None] * (sum(df.npartitions for df in dfs) + 1) return stack_partitions( dfs, divisions, join=join, ignore_order=ignore_order, **kwargs ) def _contains_index_name(df, columns_or_index): """ Test whether ``columns_or_index`` contains a reference to the index of ``df This is the local (non-collection) version of ``dask.core.DataFrame._contains_index_name``. """ def _is_index_level_reference(x, key): return ( x.index.name is not None and (np.isscalar(key) or isinstance(key, tuple)) and key == x.index.name and key not in getattr(x, "columns", ()) ) if isinstance(columns_or_index, list): return any(_is_index_level_reference(df, n) for n in columns_or_index) else: return _is_index_level_reference(df, columns_or_index) def _select_columns_or_index(df, columns_or_index): """ Returns a DataFrame with columns corresponding to each column or index level in columns_or_index. If included, the column corresponding to the index level is named _index. This is the local (non-collection) version of ``dask.core.DataFrame._select_columns_or_index``. """ def _is_column_label_reference(df, key): return (np.isscalar(key) or isinstance(key, tuple)) and key in df.columns # Ensure columns_or_index is a list columns_or_index = ( columns_or_index if isinstance(columns_or_index, list) else [columns_or_index] ) column_names = [n for n in columns_or_index if _is_column_label_reference(df, n)] selected_df = df[column_names] if _contains_index_name(df, columns_or_index): # Index name was included selected_df = selected_df.assign(_index=df.index) return selected_df def _split_partition(df, on, nsplits): """ Split-by-hash a DataFrame into `nsplits` groups. Hashing will be performed on the columns or index specified by `on`. """ if isinstance(on, bytes): on = pickle.loads(on) if isinstance(on, str) or pd.api.types.is_list_like(on): # If `on` is a column name or list of column names, we # can hash/split by those columns. on = [on] if isinstance(on, str) else list(on) nset = set(on) if nset.intersection(set(df.columns)) == nset: ind = hash_object_dispatch(df[on], index=False) ind = ind % nsplits return group_split_dispatch(df, ind.values, nsplits, ignore_index=False) # We are not joining (purely) on columns. Need to # add a "_partitions" column to perform the split. if not isinstance(on, _Frame): on = _select_columns_or_index(df, on) partitions = partitioning_index(on, nsplits) df2 = df.assign(_partitions=partitions) return shuffle_group( df2, ["_partitions"], 0, nsplits, nsplits, False, nsplits, ) def _concat_wrapper(dfs): """Concat and remove temporary "_partitions" column""" df = _concat(dfs, False) if "_partitions" in df.columns: del df["_partitions"] return df def _merge_chunk_wrapper(*args, **kwargs): return merge_chunk( *args, **{ k: pickle.loads(v) if isinstance(v, bytes) else v for k, v in kwargs.items() }, ) def broadcast_join( lhs, left_on, rhs, right_on, how="inner", npartitions=None, suffixes=("_x", "_y"), shuffle=None, indicator=False, parts_out=None, ): """Join two DataFrames on particular columns by broadcasting This broadcasts the partitions of the smaller DataFrame to each partition of the larger DataFrame, joins each partition pair, and then concatenates the new data for each output partition. """ if npartitions: # Repartition the larger collection before the merge if lhs.npartitions < rhs.npartitions: rhs = rhs.repartition(npartitions=npartitions) else: lhs = lhs.repartition(npartitions=npartitions) if how not in ("inner", "left", "right"): # Broadcast algorithm cannot handle an "outer" join raise ValueError( "Only 'inner', 'left' and 'right' broadcast joins are supported." ) if how == "left" and lhs.npartitions < rhs.npartitions: # Must broadcast rhs for a "left" broadcast join raise ValueError("'left' broadcast join requires rhs broadcast.") if how == "right" and rhs.npartitions <= lhs.npartitions: # Must broadcast lhs for a "right" broadcast join raise ValueError("'right' broadcast join requires lhs broadcast.") # TODO: It *may* be beneficial to perform the hash # split for "inner" join as well (even if it is not # technically needed for correctness). More testing # is needed here. if how != "inner": # Shuffle to-be-broadcasted side by hash. This # means that we will need to perform a local # shuffle and split on each partition of the # "other" collection (with the same hashing # approach) to ensure the correct rows are # joined by `merge_chunk`. The local hash and # split of lhs is in `_split_partition`. if lhs.npartitions < rhs.npartitions: lhs2 = shuffle_func( lhs, left_on, shuffle="tasks", ) lhs_name = lhs2._name lhs_dep = lhs2 rhs_name = rhs._name rhs_dep = rhs else: rhs2 = shuffle_func( rhs, right_on, shuffle="tasks", ) lhs_name = lhs._name lhs_dep = lhs rhs_name = rhs2._name rhs_dep = rhs2 else: lhs_name = lhs._name lhs_dep = lhs rhs_name = rhs._name rhs_dep = rhs if isinstance(left_on, Index): left_on = None left_index = True else: left_index = False if isinstance(right_on, Index): right_on = None right_index = True else: right_index = False merge_kwargs = dict( how=how, left_on=left_on, right_on=right_on, left_index=left_index, right_index=right_index, suffixes=suffixes, indicator=indicator, ) # dummy result meta = lhs._meta_nonempty.merge(rhs._meta_nonempty, **merge_kwargs) merge_kwargs["empty_index_dtype"] = meta.index.dtype merge_kwargs["categorical_columns"] = meta.select_dtypes(include="category").columns # Assume the output partitions/divisions # should correspond to the collection that # is NOT broadcasted. if lhs.npartitions < rhs.npartitions: npartitions = rhs.npartitions divisions = rhs.divisions _index_names = set(rhs._meta_nonempty.index.names) else: npartitions = lhs.npartitions divisions = lhs.divisions _index_names = set(lhs._meta_nonempty.index.names) # Cannot preserve divisions if the index is lost if _index_names != set(meta.index.names): divisions = [None] * (npartitions + 1) token = tokenize(lhs, rhs, npartitions, **merge_kwargs) name = "bcast-join-" + token broadcast_join_layer = BroadcastJoinLayer( name, npartitions, lhs_name, lhs.npartitions, rhs_name, rhs.npartitions, parts_out=parts_out, **merge_kwargs, ) graph = HighLevelGraph.from_collections( name, broadcast_join_layer, dependencies=[lhs_dep, rhs_dep], ) return new_dd_object(graph, name, meta, divisions) def _recursive_pairwise_outer_join( dataframes_to_merge, on, lsuffix, rsuffix, npartitions, shuffle ): """ Schedule the merging of a list of dataframes in a pairwise method. This is a recursive function that results in a much more efficient scheduling of merges than a simple loop from: [A] [B] [C] [D] -> [AB] [C] [D] -> [ABC] [D] -> [ABCD] to: [A] [B] [C] [D] -> [AB] [CD] -> [ABCD] Note that either way, n-1 merges are still required, but using a pairwise reduction it can be completed in parallel. :param dataframes_to_merge: A list of Dask dataframes to be merged together on their index :return: A single Dask Dataframe, comprised of the pairwise-merges of all provided dataframes """ number_of_dataframes_to_merge = len(dataframes_to_merge) merge_options = { "on": on, "lsuffix": lsuffix, "rsuffix": rsuffix, "npartitions": npartitions, "shuffle": shuffle, } # Base case 1: just return the provided dataframe and merge with `left` if number_of_dataframes_to_merge == 1: return dataframes_to_merge[0] # Base case 2: merge the two provided dataframe to be merged with `left` if number_of_dataframes_to_merge == 2: merged_ddf = dataframes_to_merge[0].join( dataframes_to_merge[1], how="outer", **merge_options ) return merged_ddf # Recursive case: split the list of dfs into two ~even sizes and continue down else: middle_index = number_of_dataframes_to_merge // 2 merged_ddf = _recursive_pairwise_outer_join( [ _recursive_pairwise_outer_join( dataframes_to_merge[:middle_index], **merge_options ), _recursive_pairwise_outer_join( dataframes_to_merge[middle_index:], **merge_options ), ], **merge_options, ) return merged_ddf
{ "alphanum_fraction": 0.6007355715, "author": null, "avg_line_length": 33.2413793103, "converted": null, "ext": "py", "file": null, "hexsha": "169c1c76b670702cd985b6c2681318522a53ceb6", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "fc1cea9cdb2ea31348204aa51e4f6f7327a2af33", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "scharlottej13/dask", "max_forks_repo_path": "dask/dataframe/multi.py", "max_issues_count": 1, "max_issues_repo_head_hexsha": "56eeb06103efbf36cc73e9405bcec42a5b92515a", "max_issues_repo_issues_event_max_datetime": "2021-12-02T20:42:37.000Z", "max_issues_repo_issues_event_min_datetime": "2021-12-01T20:16:41.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "SultanOrazbayev/dask", "max_issues_repo_path": "dask/dataframe/multi.py", "max_line_length": 120, "max_stars_count": null, "max_stars_repo_head_hexsha": "56eeb06103efbf36cc73e9405bcec42a5b92515a", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "SultanOrazbayev/dask", "max_stars_repo_path": "dask/dataframe/multi.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 12369, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 53020 }
In order to give a better sense of the approach to reliability analysis and optimization presented in \sref{chaos-reliability-analysis} and \sref{chaos-optimization}, we consider a concrete application, meaning that we specify the uncertain parameters and discuss the accompanying computations. This application is also utilized for the quantitative evaluation of our technique presented in the next section, \sref{chaos-optimization-results}. \subsection{\problemtitle} Assume that the structure of the reliability model $R(\cdot | \vg)$ of the system at hand is the one given in \eref{reliability-model} where each individual reliability function $R_i(\cdot | \vg_i)$ is the one shown in \eref{weibull-reliability} with its own parameters $\scale_i$ and $\shape_i$. During each iteration, the temperature of processing element~$i$ exhibits \nk{i} cycles. Each cycle generally has different characteristics and hence causes a different amount of damage to the processing element. This aspect is accounted for by adjusting $\scale_i$ as shown in \eref{thermal-cycling-scale}. The shape parameter $\shape_i$ is known to be indifferent to temperature \cite{chang2006}. For simplicity, assume that $\shape_i$ does not depend on process parameters either, and that $\shape_i = \shape$ for $i = \range{1}{\np}$. Under the above assumptions, \rref{weibull-homogeneity} applies, and the lifetime $\life: \Omega \to \real$ of the system has a Weibull distribution as follows: \[ \life | (\scale, \shape) \sim \mathrm{Weibull}(\scale, \shape) \] where $\scale$ is the one given in \rref{weibull-homogeneity} combined with \eref{thermal-cycling-scale}. Even though the reliability model has two parameters, only one of them is uncertain to the designer, namely $\scale$. Therefore, we treat the model as if it was parameterized only by $\scale$. The shape parameter $\shape$ is assumed to be implicitly given. In the case of reliability analysis under process variation without any accompanying exploration of the design space, one can proceed to constructing a \ac{PC} expansion of $\scale$. Having obtained this lightweight surrogate, the reliability of the system can be studied from various perspectives. In the current scenario, however, the quantity of interest \g is the one given in \eref{chaos-optimization-quantity}, since it allows for evaluating the objective function and constraints defined in \eref{chaos-optimization-objective} and \eref{chaos-optimization-constraints}, respectively. In \eref{chaos-optimization-quantity}, the component denoted by \life stands for the parameterization of the reliability model; consequently, it is $\scale$ in the illustrative application developed in this section. Let us now turn our attention to the uncertain parameters \vu of the problem being addressed. We focus on two crucial process parameters: the effective channel length and gate oxide thickness. Each processing element is then assigned two random variables corresponding to the two process parameters, which means that $\nu = 2 \np$ in the current example; see also \sref{chaos-problem}. \begin{remark} The variability in a process parameter at a spatial location can be modeled as a composition of several parts---such as inter-lot, inter-wafer, inter-die, and intra-die variations---which is demonstrated in \sref{chaos-transient-application}. In this section, we illustrate a different approach. From a mathematical perspective, it is sufficient to consider only one random variable per location with an adequate distribution and correlations with respect to the other locations. \end{remark} Based on \sref{chaos-formulation}, the parameters \vu are assumed to be given as a set of marginal distributions and a correlation matrix denoted by $\set{F_i}_{i = 1}^\nu$ and $\correlation{\vu}$, respectively. Note that the number of distinct marginals is only two, since \np components of \vu correspond to the same process parameter. Both process parameters, the effective channel length and gate oxide thickness, correspond to Euclidean distances; they take values on bounded intervals of the positive half of the real line. Consequently, similarly to \sref{chaos-transient-application}, we model the two process parameters using the four-parameter family of beta distributions shown in \eref{beta-distribution}. Without loss of generality, the parameters are assumed to be independent of each other, and the correlations between those elements of \vu that correspond to the same process parameter are assumed to be given by the correlation function shown in \eref{bayes-correlation}. The process parameters manifest themselves in the calculations associated with the power model shown in \eref{chaos-power-model-bulk} through static power. Analogously to \sref{chaos-transient-application}, the modeling here is based on \up{SPICE} simulations of a series of \up{CMOS} invertors. The invertors are taken from the 45-nm open cell library by NanGate \cite{nangate} and configured according to the 45-nm \up{PTM} \up{HP} model \cite{ptm}. The simulations are performed on a fine-grained and sufficiently broad three-dimensional grid comprising the effective channel length, gate oxide thickness, and temperature; the results are tabulated. An interpolation algorithm is subsequently employed whenever static power is to be evaluated at a particular point within the range of the grid. The output of this model is scaled up to account for about 40\% of the total power consumption \cite{liu2007}. Regarding temperature, the thermal \up{RC} circuit utilized for dynamic steady-state analysis is constructed by virtue of HotSpot \cite{skadron2003} as described in \sref{temperature-model}. At this point, the two outputs of Stage~1 are now specified. \subsection{Probability Transformation} At Stage~2 in \fref{chaos-overview}, the uncertain parameters \vu are transformed into a vector of independent random variables \vz via a suitable transformation $\transform$. Specifically, we use the one given in \eref{probability-transformation}, which also includes model order reduction. Unlike \sref{chaos-transient-application}, in this section, we let \vz obey the standard Gaussian distribution and, therefore, tailor $\transform$ accordingly; see \xref{probability-transformation}. \subsection{Surrogate Construction} Since the auxiliary variables $\vz = (\z_i)_{i = 1}^\nz$ are Gaussian, the polynomial basis considered at Stage~3 is to be composed of Hermite polynomials, which is the exact scenario described in \xref{polynomial-chaos}. The variables also tell us how to approach numerical integration needed for evaluation of the coefficients of \ac{PC} expansions: since we are interested in integrals with respect to the standard Gaussian measure, Gauss--Hermite quadratures \cite{maitre2010} are worth considering. These quadratures are especially efficient, since they belong to the class of Gaussian quadratures and thus inherit their properties; see \xref{numerical-integration}. Lastly, let us illustrate the Hermite basis. In the case of working with only one standard Gaussian variable ($\nz = 1$), a second-level \ac{PC} expansion ($\lc = 2$) of a three-dimensional quantity of interest \vg is as follows: \[ \chaos{1}{2}{\vg} = \hat{\vg}_{(0)} \psi_{(0)} + \hat{\vg}_{(1)} \psi_{(1)} + \hat{\vg}_{(2)} \psi_{(2)} \] where $\set{\hat{\vg}_{\vi}} \subset \real^3$, \begin{align*} & \psi_{(0)}(\vz) = 1, \\ & \psi_{(1)}(\vz) = \z_1, \text{ and} \\ & \psi_{(2)}(\vz) = \z_1^2 - 1. \end{align*} At Stage~4, the expansion is post-processed as described in \sref{chaos-optimization}.
{ "alphanum_fraction": 0.7832131148, "author": null, "avg_line_length": 56.4814814815, "converted": null, "ext": "tex", "file": null, "hexsha": "e99f311050dff10e93558ec47c98125975d34d16", "include": null, "lang": "TeX", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "95a7e2ee7664b94156906322610555e36e53cfe0", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "IvanUkhov/thesis", "max_forks_repo_path": "include/uncertainty/process/development/optimization-application.tex", "max_issues_count": null, "max_issues_repo_head_hexsha": "95a7e2ee7664b94156906322610555e36e53cfe0", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "IvanUkhov/thesis", "max_issues_repo_path": "include/uncertainty/process/development/optimization-application.tex", "max_line_length": 80, "max_stars_count": null, "max_stars_repo_head_hexsha": "95a7e2ee7664b94156906322610555e36e53cfe0", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "IvanUkhov/thesis", "max_stars_repo_path": "include/uncertainty/process/development/optimization-application.tex", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1870, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 7625 }
import os import pytest import musdb import simplejson as json import museval import numpy as np json_path = os.path.join( os.path.dirname(os.path.realpath(__file__)), 'data/Music Delta - Rock.json', ) @pytest.fixture() def mus(): return musdb.DB(root_dir='data/MUS-STEMS-SAMPLE', is_wav=True) def test_evaluate_mus_dir(mus): museval.eval_mus_dir( dataset=mus, # instance of musdb estimates_dir='data/EST', # path to estimate folder output_dir='data/EST_scores_mus_dir', # set a folder to write eval ) def test_eval_dir(mus): with pytest.raises(ValueError): museval.eval_dir( reference_dir='data/EST', # path to estimate folder estimates_dir='data/EST', # set a folder to write eval json files ) def test_estimate_and_evaluate(mus): # return any number of targets with open(json_path) as json_file: ref = json.loads(json_file.read()) print(os.path.basename(json_path)) track = mus.load_mus_tracks( tracknames=[os.path.splitext(os.path.basename(json_path))[0]] )[0] np.random.seed(0) random_voc = np.random.random(track.audio.shape) random_acc = np.random.random(track.audio.shape) # create a silly regression test estimates = { 'vocals': random_voc, 'accompaniment': random_acc } scores = museval.eval_mus_track( track, estimates ) assert scores.validate() is None with open( os.path.join('.', track.name) + '.json', 'w+' ) as f: f.write(scores.json) scores = json.loads(scores.json) for target in ref['targets']: for metric in ['SDR', 'SIR', 'SAR', 'ISR']: ref = np.array([d['metrics'][metric] for d in target['frames']]) idx = [t['name'] for t in scores['targets']].index(target['name']) est = np.array( [ d['metrics'][metric] for d in scores['targets'][idx]['frames'] ] ) assert np.allclose(ref, est)
{ "alphanum_fraction": 0.6033573141, "author": null, "avg_line_length": 25.4268292683, "converted": null, "ext": "py", "file": null, "hexsha": "b89de0cc8403e81c3eb63f62b1bba82b7784bbe9", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "cec9b1aae45400a04a426ab214185efc89c0c563", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "pseeth/sigsep-mus-eval", "max_forks_repo_path": "tests/test_regression.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "cec9b1aae45400a04a426ab214185efc89c0c563", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "pseeth/sigsep-mus-eval", "max_issues_repo_path": "tests/test_regression.py", "max_line_length": 78, "max_stars_count": null, "max_stars_repo_head_hexsha": "cec9b1aae45400a04a426ab214185efc89c0c563", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "pseeth/sigsep-mus-eval", "max_stars_repo_path": "tests/test_regression.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 511, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 2085 }
// Boost.Geometry (aka GGL, Generic Geometry Library) // Unit Test // Copyright (c) 2015, Oracle and/or its affiliates. // Contributed and/or modified by Menelaos Karavelas, on behalf of Oracle // Licensed under the Boost Software License version 1.0. // http://www.boost.org/users/license.html #ifndef BOOST_GEOMETRY_TEST_ENVELOPE_EXPAND_ON_SPHEROID_HPP #define BOOST_GEOMETRY_TEST_ENVELOPE_EXPAND_ON_SPHEROID_HPP #include <cmath> #include <cstddef> #include <algorithm> #include <boost/type_traits/is_same.hpp> #include <boost/geometry/core/access.hpp> #include <boost/geometry/core/coordinate_dimension.hpp> #include <boost/geometry/core/cs.hpp> #include <boost/geometry/util/condition.hpp> #include <boost/geometry/util/math.hpp> #include <boost/geometry/views/detail/indexed_point_view.hpp> #include <boost/geometry/algorithms/assign.hpp> template <typename Units> char const* units2string() { if (BOOST_GEOMETRY_CONDITION((boost::is_same<Units, bg::degree>::value))) { return "degrees"; } return "radians"; } template <typename CoordinateSystem> struct other_system_info {}; template <> struct other_system_info<bg::cs::spherical_equatorial<bg::radian> > { typedef bg::degree units; typedef bg::cs::spherical_equatorial<units> type; template <typename T> static inline T convert(T const& value) { return value * bg::math::r2d<T>(); } }; template <> struct other_system_info<bg::cs::spherical_equatorial<bg::degree> > { typedef bg::radian units; typedef bg::cs::spherical_equatorial<units> type; template <typename T> static inline T convert(T const& value) { return value * bg::math::d2r<T>(); } }; template <> struct other_system_info<bg::cs::geographic<bg::radian> > { typedef bg::degree units; typedef bg::cs::geographic<units> type; template <typename T> static inline T convert(T const& value) { return value * bg::math::r2d<T>(); } }; template <> struct other_system_info<bg::cs::geographic<bg::degree> > { typedef bg::radian units; typedef bg::cs::geographic<units> type; template <typename T> static inline T convert(T const& value) { return value * bg::math::d2r<T>(); } }; class equals_with_tolerance { private: double m_tolerence; template <typename T> static inline T const& get_max(T const& a, T const& b, T const& c) { return (std::max)((std::max)(a, b), c); } template <typename T> static inline bool check_close(T const& a, T const& b, double tol) { return (a == b) || (std::abs(a - b) <= tol * get_max(std::abs(a), std::abs(b), 1.0)); } public: equals_with_tolerance(double tolerence) : m_tolerence(tolerence) {} template <typename T> inline bool operator()(T const& value1, T const& value2) const { return check_close(value1, value2, m_tolerence); } }; template < typename Box1, typename Box2 = Box1, std::size_t DimensionCount = bg::dimension<Box1>::value > struct box_equals { static inline bool apply(Box1 const& box1, Box2 const& box2, double tol) { equals_with_tolerance equals(tol); #ifndef BOOST_GEOMETRY_TEST_ENABLE_FAILING // check latitude with tolerance when necessary return bg::math::equals(bg::get<0, 0>(box1), bg::get<0, 0>(box2)) && (bg::get<0, 1>(box1) < 0 ? equals(bg::get<0, 1>(box1), bg::get<0, 1>(box2)) : bg::math::equals(bg::get<0, 1>(box1), bg::get<0, 1>(box2))) && bg::math::equals(bg::get<1, 0>(box1), bg::get<1, 0>(box2)) && (bg::get<1, 1>(box1) > 0 ? equals(bg::get<1, 1>(box1), bg::get<1, 1>(box2)) : bg::math::equals(bg::get<1, 1>(box1), bg::get<1, 1>(box2))); #else // check latitude with tolerance when necessary return bg::get<0, 0>(box1) == bg::get<0, 0>(box2) && (bg::get<0, 1>(box1) < 0 ? equals(bg::get<0, 1>(box1), bg::get<0, 1>(box2)) : bg::get<0, 1>(box1) == bg::get<0, 1>(box2)) && bg::get<1, 0>(box1) == bg::get<1, 0>(box2) && (bg::get<1, 1>(box1) > 0 ? equals(bg::get<1, 1>(box1), bg::get<1, 1>(box2)) : bg::get<1, 1>(box1) == bg::get<1, 1>(box2)); #endif } }; template <typename Box1, typename Box2> struct box_equals<Box1, Box2, 3> { static inline bool apply(Box1 const& box1, Box2 const& box2, double tol) { #ifndef BOOST_GEOMETRY_TEST_ENABLE_FAILING equals_with_tolerance equals(tol); return box_equals<Box1, Box2, 2>::apply(box1, box2, tol) && equals(bg::get<0, 2>(box1), bg::get<0, 2>(box2)) && equals(bg::get<1, 2>(box1), bg::get<1, 2>(box2)); #else return box_equals<Box1, Box2, 2>::apply(box1, box2, tol) && bg::get<0, 2>(box1) == bg::get<0, 2>(box2) && bg::get<1, 2>(box1) == bg::get<1, 2>(box2); #endif } }; template <typename Box, std::size_t Dimension = bg::dimension<Box>::value> struct initialize_box { static inline void apply(Box& box, double lon_min, double lat_min, double, double lon_max, double lat_max, double) { bg::detail::indexed_point_view<Box, bg::min_corner> p_min(box); bg::detail::indexed_point_view<Box, bg::max_corner> p_max(box); bg::assign_values(p_min, lon_min, lat_min); bg::assign_values(p_max, lon_max, lat_max); } }; template <typename Box> struct initialize_box<Box, 3> { static inline void apply(Box& box, double lon_min, double lat_min, double height_min, double lon_max, double lat_max, double height_max) { bg::detail::indexed_point_view<Box, bg::min_corner> p_min(box); bg::detail::indexed_point_view<Box, bg::max_corner> p_max(box); bg::assign_values(p_min, lon_min, lat_min, height_min); bg::assign_values(p_max, lon_max, lat_max, height_max); } }; #endif // BOOST_GEOMETRY_TEST_ENVELOPE_EXPAND_ON_SPHEROID_HPP
{ "alphanum_fraction": 0.6170454545, "author": null, "avg_line_length": 28.5185185185, "converted": null, "ext": "hpp", "file": null, "hexsha": "14066cd472543364ca9b0055b406937a792d36fb", "include": null, "lang": "C++", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 2, "max_forks_repo_forks_event_max_datetime": "2016-08-11T20:31:46.000Z", "max_forks_repo_forks_event_min_datetime": "2016-07-30T10:17:12.000Z", "max_forks_repo_head_hexsha": "66ea1fd4946668192e3f0d1060f0844f324ad7b8", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "lipper/arangodb", "max_forks_repo_path": "3rdParty/boost/1.62.0/libs/geometry/test/algorithms/envelope_expand/test_envelope_expand_on_spheroid.hpp", "max_issues_count": 49, "max_issues_repo_head_hexsha": "66ea1fd4946668192e3f0d1060f0844f324ad7b8", "max_issues_repo_issues_event_max_datetime": "2019-05-05T04:59:26.000Z", "max_issues_repo_issues_event_min_datetime": "2016-02-29T17:59:52.000Z", "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "lipper/arangodb", "max_issues_repo_path": "3rdParty/boost/1.62.0/libs/geometry/test/algorithms/envelope_expand/test_envelope_expand_on_spheroid.hpp", "max_line_length": 81, "max_stars_count": 18, "max_stars_repo_head_hexsha": "6a4f462fa209010cd064f99e63d85ce1d432c500", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "sita1999/arangodb", "max_stars_repo_path": "3rdParty/boost/1.62.0/libs/geometry/test/algorithms/envelope_expand/test_envelope_expand_on_spheroid.hpp", "max_stars_repo_stars_event_max_datetime": "2021-12-31T11:06:25.000Z", "max_stars_repo_stars_event_min_datetime": "2016-03-04T15:44:24.000Z", "num_tokens": 1775, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 6160 }
#Djnago stuff from .models import * from django.shortcuts import render,redirect from django.http import HttpResponse,JsonResponse from django.core.serializers import serialize import time #helper functions from .utils import traverse,convertGraph #run binaries import subprocess from subprocess import PIPE, run #networkx converters from networkx.readwrite import json_graph from networkx.readwrite import parse_gml,from_graph6_bytes,to_graph6_bytes from networkx.exception import NetworkXException import networkx as nx #json module import json #path building from pathlib import Path import os def graphPage(request): if request.method == 'POST': #get data from the form which is on home page if ("file" in request.FILES): try: data = request.FILES['file'].read().decode("utf-8") except: return render(request,"graph_visualization/error.html") else: data=request.POST.get("text") #some networkx converters dont handle strings(gexf,edge list,) so we need to create a temp file from #which we can read with open('temp.txt', 'w') as f: f.write(data) renderingOption=request.POST.get("radio") context=dict() graphData=convertGraph(data) context['data']=graphData context["commands"]= Command.objects.all() #render 2d or 3d page if graphData == "error": return render(request,"graph_visualization/error.html",context) elif renderingOption == "2D": return render(request,"graph_visualization/graph.html",context) else: return render(request,"graph_visualization/graph3d.html",context) elif request.method== "GET" : context=dict() g6Str=request.GET.get("g6") graphData=convertGraph(g6Str) context['data']=graphData context["commands"]= Command.objects.all() return render(request,"graph_visualization/graph.html",context) def graphPage3D(request): context={} return render(request,"graph_visualization/graph3d.html",context) def home(request): context=dict() context["home"]=1 return render(request, 'graph_visualization/home.html',context) def About(request): context=dict() context["home"]=1 return render(request, 'graph_visualization/about.html',context) def Manual(request): context=dict() context["home"]=1 return render(request, 'graph_visualization/manual.html',context) def JsonEditor(request): return render(request, 'graph_visualization/jsonEditor.html',) def LoadNewContent(request): if request.method == 'POST': if ("file" in request.FILES): try: data = request.FILES['file'].read().decode("utf-8") except: return render(request,"graph_visualization/error.html") with open('temp.txt', 'w') as f: f.write(data) print("loadcont") graphData=convertGraph(data) return JsonResponse({"data":graphData}) def exportGML(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" return JsonResponse({"data":response}) def exportG6(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) response=to_graph6_bytes(G) response=response.decode().replace(">>graph6<<","").strip() return JsonResponse({"data":response}) def exportGraphML(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) attrs = set() #get attributes in nested structures for node in G.nodes: for attrib in G.nodes[node]: if type(G.nodes[node][attrib]) == dict: for key in G.nodes[node][attrib].keys(): attrs.add(key) #G.nodes[node][attrib]="" print(attrs) for attr in attrs: if isinstance(attr, int): var =int() elif isinstance(attr,str): var = str() nx.set_node_attributes(G, var, attr) for node in G.nodes: for attrib in G.nodes[node]: if type(G.nodes[node][attrib]) == dict: #graphics dict print(G.nodes[node][attrib]) d=G.nodes[node][attrib] for key, value in d.items(): print(G.nodes[node][attrib]) G.nodes[node][key]=value G.nodes[node][attrib]="" print(G) for edge in G.edges: for attrib in G.edges[edge]: if type(G.edges[edge][attrib]) == dict: G.edges[edge][attrib]="" response="" linefeed = chr(10) s = linefeed.join(nx.generate_graphml(G)) for line in nx.generate_graphml(G): response+=line+"\n" return JsonResponse({"data":response}) def exportGexf(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) response="" for line in nx.generate_gexf(G): response+=line+"\n" return JsonResponse({"data":response}) def exportEdgelist(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) response="" for line in nx.generate_edgelist(G, data=False): response+="\n"+line return JsonResponse({"data":response.strip("\n")}) def exportAdjacencyList(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) response="" for line in nx.generate_adjlist(G): response+=line+"\n" return JsonResponse({"data":response}) def exportSparse6(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) response=nx.to_sparse6_bytes(G).decode() return JsonResponse({"data":response}) def exportDot(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" G=parse_gml(response,label="id",destringizer=None) response=nx.nx_pydot.to_pydot(G).to_string() #output_graphviz_svg = response.create_svg() #print(output_graphviz_svg) return JsonResponse({"data":response}) #binary file view def modifyAPI(request): if request.method == 'POST': json_data = json.loads(request.body) nodes = json_data["nodes"] links = json_data["links"] cmd = json_data["cmd"] parameters =json_data["parameters"] print(parameters) response="graph [\n" for node in nodes: response+="node [\n" response+=traverse(node) response+="]\n" for link in links: response+="edge [\n" response+=traverse(link) response+="]\n" #end of graph response+="]" #search for the specific binary name in the DB binary = Command.objects.get(name=cmd) #save the actual terminal command in a variable cmd=binary.terminal_command #add the params if they exist for key,value in parameters.items(): if (key!=""): cmd+=" -"+key if(value!=""): cmd+=" "+value print(cmd) #start timer to see how long it takes to run the binary start_time = time.time() with open('testfile.gml', 'w') as f: f.write(response) popen = subprocess.Popen(cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE,shell=True) output,error =popen.communicate() print(error) print("--- %s seconds ---" % (time.time() - start_time)) with open("testfile-out.graphml", 'r') as f: gml=f.read() gml.replace('Creator "ogdf::GraphIO::writeGML"',"") output=convertGraph(gml) return HttpResponse(output) def getCommandParameter(request): if request.method == 'POST': name = json.loads(request.body) #search for the specific binary name in the DB binary = Command.objects.get(name=name) #get parameters from DB as an object parameters= CommandParameter.objects.filter(command=binary.id) data = list(parameters.values()) #jsonString = json.dumps(data) return JsonResponse(data,safe=False) def getSampleGraph(request): if request.method == 'POST': name = json.loads(request.body) with open("sample_graphs/"+ name +".txt", 'r') as f: graphString=f.read() return JsonResponse({"data":graphString})
{ "alphanum_fraction": 0.551655365, "author": null, "avg_line_length": 28.6187214612, "converted": null, "ext": "py", "file": null, "hexsha": "25cf1d93e52afa109a0fa9687cebe1680ff7bd2e", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-07-18T12:49:17.000Z", "max_forks_repo_forks_event_min_datetime": "2021-07-18T12:49:17.000Z", "max_forks_repo_head_hexsha": "8eab2fa79251dd78ded46f440269600f87de115a", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "Skazu14/grapp", "max_forks_repo_path": "main/graph_visualization/views.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "8eab2fa79251dd78ded46f440269600f87de115a", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "Skazu14/grapp", "max_issues_repo_path": "main/graph_visualization/views.py", "max_line_length": 108, "max_stars_count": null, "max_stars_repo_head_hexsha": "8eab2fa79251dd78ded46f440269600f87de115a", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "Skazu14/grapp", "max_stars_repo_path": "main/graph_visualization/views.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2669, "path": null, "reason": "import networkx,from networkx", "repo": null, "save_path": null, "sha": null, "size": 12535 }
from datetime import datetime, timezone import numpy as np import pytest import astropy.units as u from stixcore.time import SCETime, SCETimeDelta, SCETimeRange from stixcore.time.datetime import MAX_COARSE, MAX_FINE def test_time_init(): t1 = SCETime(0, 0) t2 = SCETime.from_float(0*u.s) t3 = SCETime(t1) assert t1 == t2 assert t2 == t3 assert t1 == t3 with pytest.raises(ValueError, match=r'Coarse time must be in range.*'): SCETime(-1, 0) with pytest.raises(ValueError, match=r'Fine time must be in range.*'): SCETime(0, -1) with pytest.raises(ValueError): _ = SCETime(2 ** 44-1, 0) with pytest.raises(ValueError): SCETime(0, 2**16+1) with pytest.raises(ValueError): SCETime(0.0, 0) def test_time_to_datetime(): dt = SCETime(coarse=0, fine=0) assert dt.to_datetime() == datetime(2000, 1, 1, 0, tzinfo=timezone.utc) def test_time_as_float(): dt = SCETime(coarse=1, fine=0) assert dt.as_float() == 1.0 * u.s def test_time_from_float(): dt = SCETime.from_float(1 * u.s) assert dt == SCETime(coarse=1, fine=0) def test_time_to_str(): dt = SCETime(coarse=123, fine=45) assert dt == SCETime.from_string(str(dt)) def test_time_add(): t1 = SCETime(123, 456) with pytest.raises(TypeError, match=r'Only Quantities and SCETimeDelta.*'): _ = t1 + SCETime(0, 1) with pytest.raises(ValueError, match=r'.*are not convertible'): _ = t1 + (1*u.m) # test right add t2 = t1 + (1 + 1/MAX_FINE) * u.s # test left add t3 = (1 + 1/MAX_FINE) * u.s + t1 assert t2 == t3 assert t2.coarse == 124 assert t2.fine == 457 def test_time_sub(): t1 = SCETime(123, 456) with pytest.raises(TypeError, match=r'Only quantities, SCETime and SCETimeDelt.*'): _ = t1 - 1 with pytest.raises(ValueError, match=r'.*are not convertible'): _ = t1 + (1*u.m) # test sub t2 = t1 - (1 + 1/MAX_FINE) * u.s assert t2.coarse == 122 assert t2.fine == 455 # test rsub with pytest.raises(TypeError, match=r'unsupported operand.*'): t2 = (1 + 1/MAX_FINE) * u.s - t1 # Test subtract to times dt = t1 - t2 assert isinstance(dt, SCETimeDelta) assert dt.coarse == 1 assert dt.fine == 1 dt = t2 - t1 assert isinstance(dt, SCETimeDelta) assert dt.coarse == -1 assert dt.fine == -1 # Test subtract deltatime t3 = t2 - dt assert isinstance(t3, SCETime) assert t3.coarse == 123 assert t3.fine == 456 # Can't subtract time from a delta time with pytest.raises(TypeError, match=r'Unsupported operation for ' r'types SCETimeDelta and SCETime'): _ = dt - t1 def test_time_eq(): t1 = SCETime(123, 456) t2 = SCETime.from_float((123 + 456/MAX_FINE)*u.s) t3 = SCETime(765, 432) assert t1 == t2 assert t1 != t3 def test_time_broadcast(): t = SCETime(0, 0) t1 = t + np.arange(100) * u.s t2 = SCETime(np.arange(100, dtype=np.int), 0) t3 = SCETime(0, np.arange(100, dtype=np.int)) assert t1.shape == (100,) assert t2.shape == (100,) assert t3.shape == (100,) def test_time_lt(): dt = SCETime(coarse=123, fine=45) dt2 = SCETime(coarse=124, fine=45) assert dt < dt2 assert dt <= dt2 assert dt2 > dt assert dt2 >= dt assert dt2 is not dt assert dt2 == SCETime.from_string(str(dt2)) def test_time_minmax(): assert SCETime.min_time() == SCETime(coarse=0, fine=0) assert SCETime.max_time() == SCETime(coarse=MAX_COARSE, fine=MAX_FINE) # TODO enable after https://github.com/i4Ds/STIXCore/issues/102 # assert SCETime.min_time() - SCETime(coarse=0, fine=1) == SCETime.min_time() with pytest.raises(ValueError, match=r'Coarse time must be in range.*'): m = SCETime.max_time() dt = SCETimeDelta(0, 1) nm = m + dt print(nm) def test_timedelta_init(): dt1 = SCETimeDelta(0, 0) dt2 = SCETimeDelta.from_float(0*u.s) dt3 = SCETimeDelta(dt1) assert dt1 == dt2 assert dt2 == dt3 assert dt1 == dt3 with pytest.raises(ValueError): _ = SCETimeDelta(2 ** 32 + 1, 0) with pytest.raises(ValueError): SCETime(0, 2**16+1) with pytest.raises(ValueError): SCETime(0.0, 0) def test_timedelta_as_float(): dt = SCETimeDelta(coarse=-1, fine=0) assert dt.as_float() == -1.0 * u.s def test_timedelta_from_float(): dt = SCETimeDelta.from_float(-1 * u.s) assert dt == SCETimeDelta(coarse=-1, fine=0) def test_timedelta_add(): t1 = SCETime(1, 1) dt1 = SCETimeDelta(100, 1) dt2 = SCETimeDelta(200, 2) # test time plus timedelta t1_dt1 = dt1 + t1 dt1_t1 = t1 + dt1 assert t1_dt1 == dt1_t1 assert t1_dt1.coarse == 101 assert t1_dt1.fine == 2 with pytest.raises(ValueError, match=f'.*are not convertible'): _ = dt1 + (1*u.m) # test timedelta plus timedelta/quantity dt1_dt2 = dt1 + dt2 dt1_float = dt1 + (200+2/MAX_FINE)*u.s dt2_dt1 = dt2 + dt1 float_dt2 = (100 + 1/MAX_FINE) * u.s + dt2 assert dt1_dt2 == dt2_dt1 assert dt1_float == dt1_dt2 assert float_dt2 == dt1_dt2 assert dt1_dt2.coarse == 300 assert dt1_dt2.fine == 3 def test_deltatime_sub(): t1 = SCETime(100, 2) dt1 = SCETimeDelta(100, 1) dt2 = SCETimeDelta(200, 2) with pytest.raises(TypeError, match=r'Unsupported operation for types SCETimeDelta and int'): _ = dt1 - 1 with pytest.raises(TypeError, match=r'Quantity could not be converted to SCETimeDelta'): _ = dt1 - (1*u.m) # test sub deltatimes and quantities dt1_dt2 = dt1 - dt2 dt1_float = dt1 - (200 + 2 / MAX_FINE) * u.s assert dt1_dt2 == dt1_float assert dt1_dt2.coarse == -100 assert dt1_dt2.fine == -1 dt2_dt1 = dt2 - dt1 float_dt1 = (200 + 2/MAX_FINE) * u.s - dt1 assert dt2_dt1 == float_dt1 assert dt2_dt1.coarse == 100 assert dt2_dt1.fine == 1 # test sub times with pytest.raises(TypeError, match=f'Unsupported operation for types.*'): dt1 - t1 t2 = t1 - dt1 assert t2.coarse == 0 assert t2.fine == 1 with pytest.raises(ValueError, match=r'Coarse time must be in range.*'): t1 - dt2 def test_timedelta_eq(): dt1 = SCETimeDelta(123, 456) dt2 = SCETimeDelta((123 + 456/MAX_FINE)*u.s) dt3 = SCETimeDelta(-1, -1) assert dt1 == dt2 assert dt1 != dt3 assert dt1 == (123 + 456/MAX_FINE)*u.s def test_timerange(): tr = SCETimeRange(start=SCETime(coarse=100, fine=0), end=SCETime(coarse=200, fine=0)) tp_in = SCETime(coarse=150, fine=0) tr_in = SCETimeRange(start=SCETime(coarse=150, fine=0), end=SCETime(coarse=160, fine=0)) tr_out = SCETimeRange(start=SCETime(coarse=150, fine=0), end=SCETime(coarse=250, fine=0)) tp_out = SCETime(coarse=250, fine=0) assert tp_in in tr assert tp_out not in tr assert tr_in in tr assert tr_out not in tr tr.expand(tp_out) tr.expand(tr_out) assert tp_out in tr assert tr_out in tr
{ "alphanum_fraction": 0.6251581167, "author": null, "avg_line_length": 26.8490566038, "converted": null, "ext": "py", "file": null, "hexsha": "13c9ded99596299bd7e7d929dcea175a60724c3d", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 3, "max_forks_repo_forks_event_max_datetime": "2022-01-21T07:52:51.000Z", "max_forks_repo_forks_event_min_datetime": "2020-11-09T15:05:18.000Z", "max_forks_repo_head_hexsha": "16822bbb37046f8e6c03be51909cfc91e9822cf7", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "nicHoch/STIXCore", "max_forks_repo_path": "stixcore/time/tests/test_datetime.py", "max_issues_count": 192, "max_issues_repo_head_hexsha": "16822bbb37046f8e6c03be51909cfc91e9822cf7", "max_issues_repo_issues_event_max_datetime": "2022-03-31T15:17:13.000Z", "max_issues_repo_issues_event_min_datetime": "2020-11-03T22:40:19.000Z", "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "nicHoch/STIXCore", "max_issues_repo_path": "stixcore/time/tests/test_datetime.py", "max_line_length": 97, "max_stars_count": 1, "max_stars_repo_head_hexsha": "16822bbb37046f8e6c03be51909cfc91e9822cf7", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "nicHoch/STIXCore", "max_stars_repo_path": "stixcore/time/tests/test_datetime.py", "max_stars_repo_stars_event_max_datetime": "2022-03-31T13:42:43.000Z", "max_stars_repo_stars_event_min_datetime": "2022-03-31T13:42:43.000Z", "num_tokens": 2389, "path": null, "reason": "import numpy,import astropy", "repo": null, "save_path": null, "sha": null, "size": 7115 }
import os import sys import time import pickle import random import tarfile import matplotlib import numpy as np import matplotlib.pyplot as plt import matplotlib.image as mpimg from scipy import ndimage from urllib.request import urlretrieve from sklearn.linear_model import LogisticRegression from utils import * from dataset_utils import * logger = logging.getLogger(os.path.basename(__file__)) url = 'https://commondatastorage.googleapis.com/books1000/' last_percent_reported = None data_root = '.' # Change me to store data elsewhere np.random.seed(133) pixel_depth = 255.0 # Number of levels per pixel. train_size = 200000 valid_size = 10000 test_size = 10000 def show_image_float(img, cmap=None): #plt.imshow(img, norm=matplotlib.colors.NoNorm(vmin=-0.5, vmax=0.5), cmap='gray') plt.imshow(img, vmin=-0.5, vmax=0.5, cmap='gray') plt.show() def show_image_filesystem(filename): img = mpimg.imread(filename) plt.imshow(img, cmap='gray') plt.show() def download_progress_hook(count, blockSize, totalSize): """A hook to report the progress of a download. This is mostly intended for users with slow internet connections. Reports every 5% change in download progress. """ global last_percent_reported percent = int(count * blockSize * 100 / totalSize) if last_percent_reported != percent: if percent % 5 == 0: sys.stdout.write('%s%%' % percent) sys.stdout.flush() else: sys.stdout.write('.') sys.stdout.flush() last_percent_reported = percent def maybe_download(filename, expected_bytes, force=False): """Download a file if not present, and make sure it's the right size.""" dest_filename = os.path.join(data_root, filename) if force or not os.path.exists(dest_filename): print('Attempting to download:', filename) filename, _ = urlretrieve(url + filename, dest_filename, reporthook=download_progress_hook) print('\nDownload Complete!') statinfo = os.stat(dest_filename) if statinfo.st_size == expected_bytes: print('Found and verified', dest_filename) else: raise Exception('Failed to verify ' + dest_filename + '. Can you get to it with a browser?') return dest_filename def maybe_extract(filename, force=False): root = os.path.splitext(os.path.splitext(filename)[0])[0] # remove .tar.gz if os.path.isdir(root) and not force: # You may override by setting force=True. print('%s already present - Skipping extraction of %s.' % (root, filename)) else: print('Extracting data for %s. This may take a while. Please wait.' % root) tar = tarfile.open(filename) sys.stdout.flush() tar.extractall(data_root) tar.close() data_folders = [ os.path.join(root, d) for d in sorted(os.listdir(root)) if os.path.isdir(os.path.join(root, d)) ] if len(data_folders) != NUM_CLASSES: raise Exception('Expected %d folders, one per class. Found %d instead.' % (NUM_CLASSES, len(data_folders))) print(data_folders) return data_folders def load_letter(folder, min_num_images): """Load the data for a single letter label.""" image_files = os.listdir(folder) dataset = np.ndarray(shape=(len(image_files), IMAGE_RES, IMAGE_RES), dtype=np.float32) print(folder) num_images = 0 for image in image_files: image_file = os.path.join(folder, image) try: image_data = (ndimage.imread(image_file).astype(float) - pixel_depth / 2) / pixel_depth if image_data.shape != (IMAGE_RES, IMAGE_RES): raise Exception('Unexpected image shape: %s' % str(image_data.shape)) dataset[num_images, :, :] = image_data num_images += 1 except IOError as e: print('Could not read:', image_file, ':', e, '- it\'s ok, skipping.') dataset = dataset[0:num_images, :, :] if num_images < min_num_images: raise Exception('Many fewer images than expected: %d < %d' % (num_images, min_num_images)) print('Full dataset tensor:', dataset.shape) print('Mean:', np.mean(dataset)) print('Standard deviation:', np.std(dataset)) return dataset def maybe_pickle(data_folders, min_num_images_per_class, force=False): dataset_names = [] for folder in data_folders: set_filename = folder + '.pickle' dataset_names.append(set_filename) if os.path.exists(set_filename) and not force: # You may override by setting force=True. print('%s already present - Skipping pickling.' % set_filename) else: print('Pickling %s.' % set_filename) dataset = load_letter(folder, min_num_images_per_class) try: with open(set_filename, 'wb') as f: pickle.dump(dataset, f, pickle.HIGHEST_PROTOCOL) except Exception as e: print('Unable to save data to', set_filename, ':', e) return dataset_names def make_arrays(nb_rows, img_size): if nb_rows: dataset = np.ndarray((nb_rows, img_size, img_size), dtype=np.float32) labels = np.ndarray(nb_rows, dtype=np.int32) else: dataset, labels = None, None return dataset, labels def merge_datasets(pickle_files, train_size, valid_size=0): NUM_CLASSES = len(pickle_files) valid_dataset, valid_labels = make_arrays(valid_size, IMAGE_RES) train_dataset, train_labels = make_arrays(train_size, IMAGE_RES) vsize_per_class = valid_size // NUM_CLASSES tsize_per_class = train_size // NUM_CLASSES start_v, start_t = 0, 0 end_v, end_t = vsize_per_class, tsize_per_class end_l = vsize_per_class + tsize_per_class for label, pickle_file in enumerate(pickle_files): try: with open(pickle_file, 'rb') as f: letter_set = pickle.load(f) # let's shuffle the letters to have random validation and training set np.random.shuffle(letter_set) if valid_dataset is not None: valid_letter = letter_set[:vsize_per_class, :, :] valid_dataset[start_v:end_v, :, :] = valid_letter valid_labels[start_v:end_v] = label start_v += vsize_per_class end_v += vsize_per_class train_letter = letter_set[vsize_per_class:end_l, :, :] train_dataset[start_t:end_t, :, :] = train_letter train_labels[start_t:end_t] = label start_t += tsize_per_class end_t += tsize_per_class except Exception as e: print('Unable to process data from', pickle_file, ':', e) raise return valid_dataset, valid_labels, train_dataset, train_labels def randomize(dataset, labels): permutation = np.random.permutation(labels.shape[0]) shuffled_dataset = dataset[permutation,:,:] shuffled_labels = labels[permutation] return shuffled_dataset, shuffled_labels def verify_dataset(dataset, labels): idx = random.randrange(0, labels.shape[0]) print(idx, labels[idx], chr(ord('a') + labels[idx])) plt.imshow(dataset[idx], cmap='gray') plt.show() def generate_datasets(pickle_file): train_filename = maybe_download('notMNIST_large.tar.gz', 247336696) test_filename = maybe_download('notMNIST_small.tar.gz', 8458043) print(train_filename, test_filename) train_folders = maybe_extract(train_filename) test_folders = maybe_extract(test_filename) print(train_folders, test_folders) train_datasets = maybe_pickle(train_folders, 45000) test_datasets = maybe_pickle(test_folders, 1800) print(train_datasets, test_datasets) valid_dataset, valid_labels, train_dataset, train_labels = merge_datasets( train_datasets, train_size, valid_size ) _, _, test_dataset, test_labels = merge_datasets(test_datasets, test_size) print('Training:', train_dataset.shape, train_labels.shape) print('Validation:', valid_dataset.shape, valid_labels.shape) print('Testing:', test_dataset.shape, test_labels.shape) train_dataset, train_labels = randomize(train_dataset, train_labels) test_dataset, test_labels = randomize(test_dataset, test_labels) valid_dataset, valid_labels = randomize(valid_dataset, valid_labels) if not os.path.isfile(pickle_file): try: with open(pickle_file, 'wb') as fobj: save = { 'train_dataset': train_dataset, 'train_labels': train_labels, 'valid_dataset': valid_dataset, 'valid_labels': valid_labels, 'test_dataset': test_dataset, 'test_labels': test_labels, } pickle.dump(save, fobj, pickle.HIGHEST_PROTOCOL) except Exception as e: print('Unable to save data to', pickle_file, ':', e) raise def find_duplicates(dataset1, labels1, dataset2, labels2, threshold=EPS): means = np.mean(dataset1, axis=(1,2)) indices = means.argsort() num_groups = 2000 indices = np.reshape(indices, (num_groups, indices.shape[0] // num_groups)) duplicate_indices = [] dataset2_idx = 0 for item in dataset2: belongs_to_group = 0 item_mean = item.mean() for group_idx in range(num_groups): if item_mean >= means[indices[group_idx][0]]: belongs_to_group = group_idx else: break for idx_in_group in range(len(indices[belongs_to_group])): dataset_idx = indices[belongs_to_group][idx_in_group] mse = ((item - dataset1[dataset_idx]) ** 2).mean() if mse < threshold: if labels1[dataset_idx] == labels2[dataset2_idx]: pass else: print( 'Found duplicate! Labels are different', labels1[dataset_idx], labels2[dataset2_idx], ) duplicate_indices.append(dataset2_idx) # show_image_float(item) # show_image_float(dataset1[dataset_idx]) break dataset2_idx += 1 if dataset2_idx % 100 == 0: print('Processed', dataset2_idx, 'duplicates:', len(duplicate_indices)) print('Found total of', len(duplicate_indices), 'duplicantes!') return duplicate_indices def sanitize_datasets(datasets): train_dataset, train_labels = extract_dataset(datasets, 'train') test_dataset, test_labels = extract_dataset(datasets, 'valid') valid_dataset, valid_labels = extract_dataset(datasets, 'test') duplicate_indices = find_duplicates( train_dataset, train_labels, test_dataset, test_labels, threshold=0.005, ) test_dataset_sanitized = np.delete(test_dataset, duplicate_indices, axis=0) test_labels_sanitized = np.delete(test_labels, duplicate_indices, axis=0) duplicate_indices = find_duplicates( train_dataset, train_labels, valid_dataset, valid_labels, threshold=0.005, ) valid_dataset_sanitized = np.delete(valid_dataset, duplicate_indices, axis=0) valid_labels_sanitized = np.delete(valid_labels, duplicate_indices, axis=0) pickle_file_sanitized = os.path.join(data_root, 'notMNIST_sanitized.pickle') try: with open(pickle_file_sanitized, 'wb') as fobj: save = { 'train_dataset': train_dataset, 'train_labels': train_labels, 'valid_dataset': valid_dataset_sanitized, 'valid_labels': valid_labels_sanitized, 'test_dataset': test_dataset_sanitized, 'test_labels': test_labels_sanitized, } pickle.dump(save, fobj, pickle.HIGHEST_PROTOCOL) except Exception as e: print('Unable to save data to', pickle_file_sanitized, ':', e) raise def train_logistic_classifier(x, y, x_test, y_test, num_train_samples=None): tstart = time.time() if num_train_samples is None: num_train_samples = x.shape[0] x = x[:num_train_samples] y = y[:num_train_samples] resolution = x.shape[1] x = x.reshape(num_train_samples, resolution * resolution) x_test = x_test.reshape(x_test.shape[0], resolution * resolution) classifier = LogisticRegression( C=100./num_train_samples, multi_class='multinomial', penalty='l2', solver='saga', max_iter=200, ) classifier.fit(x, y) sparsity = np.mean(classifier.coef_ == 0) * 100 score_train = classifier.score(x, y) score_test = classifier.score(x_test, y_test) print('Sparsity: %.2f%%' % sparsity) print('Train score: %.4f' % score_train) print('Test score: %.4f' % score_test) interactive = False if interactive: for i in range(10): image = x_test[i] prediction = classifier.predict(image.reshape((1, resolution * resolution)))[0] prediction_proba = classifier.predict_proba(image.reshape((1, resolution * resolution))) print('Prediction:', prediction, chr(ord('a') + prediction)) print('Prediction proba:', prediction_proba) show_image_float(image.reshape((resolution, resolution))) coef = classifier.coef_.copy() plt.figure(figsize=(10, 5)) scale = np.abs(coef).max() for i in range(10): plot = plt.subplot(2, 5, i + 1) plot.imshow( coef[i].reshape(resolution, resolution), interpolation='nearest', cmap=plt.cm.RdBu, vmin=-scale, vmax=scale, ) plot.set_xticks(()) plot.set_yticks(()) plot.set_xlabel('Class %i' % i) plt.suptitle('Classification vector for...') run_time = time.time() - tstart print('Example run in %.3f s' % run_time) plt.show() return classifier def main(): """Script entry point.""" init_logger(os.path.basename(__file__)) pickle_file = os.path.join(data_root, PICKLE_FILE) if not os.path.isfile(pickle_file): generate_datasets(pickle_file) datasets = load_datasets(pickle_file) train_dataset, train_labels = extract_dataset(datasets, 'train') test_dataset, test_labels = extract_dataset(datasets, 'valid') valid_dataset, valid_labels = extract_dataset(datasets, 'test') logger.info('%r %r', train_dataset.shape, train_labels.shape) logger.info('%r %r', test_dataset.shape, test_labels.shape) logger.info('%r %r', valid_dataset.shape, valid_labels.shape) def train(samples=None): return train_logistic_classifier( train_dataset, train_labels, test_dataset, test_labels, num_train_samples=samples, ) classifier = train(300) score_valid = classifier.score(valid_dataset.reshape(valid_dataset.shape[0], 28*28), valid_labels) print('Valid score: %.4f' % score_valid) return 0 if __name__ == '__main__': sys.exit(main())
{ "alphanum_fraction": 0.6471323336, "author": null, "avg_line_length": 37.3562653563, "converted": null, "ext": "py", "file": null, "hexsha": "1d19da61f9bf446a9638a18102d8c2594b1fa562", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "14714ee4151b798cde0a31a94ac65e08b87d0f65", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "alex-petrenko/udacity-deep-learning", "max_forks_repo_path": "assignment_01_notmnist.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "14714ee4151b798cde0a31a94ac65e08b87d0f65", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "alex-petrenko/udacity-deep-learning", "max_issues_repo_path": "assignment_01_notmnist.py", "max_line_length": 115, "max_stars_count": 7, "max_stars_repo_head_hexsha": "14714ee4151b798cde0a31a94ac65e08b87d0f65", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "alex-petrenko/udacity-deep-learning", "max_stars_repo_path": "assignment_01_notmnist.py", "max_stars_repo_stars_event_max_datetime": "2020-02-13T19:31:15.000Z", "max_stars_repo_stars_event_min_datetime": "2017-10-02T01:30:44.000Z", "num_tokens": 3406, "path": null, "reason": "import numpy,from scipy", "repo": null, "save_path": null, "sha": null, "size": 15204 }
import sys, os sys.path.append(os.getcwd()) import entities, viewer import numpy as np import importlib importlib.reload(entities) importlib.reload(viewer) cube = entities.Entity(node_color=(255, 255, 255), name="cube") cube_nodes = [(x, y, z) for x in (-75, 75) for y in (-75, 75) for z in (-75, 75)] cube.addNodes(np.array(cube_nodes)) cube.addEdges([(n, n + 4) for n in range(0, 4)]) cube.addEdges([(n, n + 1) for n in range(0, 8, 2)]) cube.addEdges([(n, n + 2) for n in (0, 1, 4, 5)]) plain = entities.Entity(node_color=(255, 255, 255), name="plane") plain_nodes = [(x, y, z) for x in (0, 150) for y in (0, 150) for z in (0,)] plain_edges = [(0, 0), (0, 1), (0, 2), (0, 3), (1, 0), (1, 1), (1, 2), (1, 3), (2, 0), (2, 1), (2, 2), (2, 3), (3, 0), (3, 1), (3, 2), (3, 3)] plain.addNodes(np.array(plain_nodes)) plain.addEdges(plain_edges) yes = viewer.Viewer(500, 500) yes.addObjects([cube, plain]) objects = ["context", "cube", "plane"] #buttons buttons = [] font_size = 20 for i in range(len(objects)): buttons.append(viewer.Button(id=i, background_color=(147, 255, 0), text_color=(255, 0, 247), top_left = (0, i*(font_size+2)), text=objects[i], font_size = font_size)) yes.addButtons(buttons) yes.run()
{ "alphanum_fraction": 0.6286885246, "author": null, "avg_line_length": 32.1052631579, "converted": null, "ext": "py", "file": null, "hexsha": "bb7adaaee71678d1d180363d9d219fc26fedfb79", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2020-08-19T17:25:22.000Z", "max_forks_repo_forks_event_min_datetime": "2020-08-19T17:25:22.000Z", "max_forks_repo_head_hexsha": "4cd94e0cf53ee06c9c31e9272572ca9656697c30", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "DarkShadow4/python", "max_forks_repo_path": "Python 3/PyGame/Matrix_based_3D/test.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "4cd94e0cf53ee06c9c31e9272572ca9656697c30", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "DarkShadow4/python", "max_issues_repo_path": "Python 3/PyGame/Matrix_based_3D/test.py", "max_line_length": 170, "max_stars_count": null, "max_stars_repo_head_hexsha": "4cd94e0cf53ee06c9c31e9272572ca9656697c30", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "DarkShadow4/python", "max_stars_repo_path": "Python 3/PyGame/Matrix_based_3D/test.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 453, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 1220 }
import numpy as np import matplotlib.pyplot as plt from IPython.display import clear_output import os import contextlib np.seterr(all='raise') class NoSignChange(Exception): def __init__(self, message='No sign change in specified bounds'): # Call the base class constructor with the parameters it needs super(NoSignChange, self).__init__(message) class BoundError(Exception): def __init__(self, message='Bounds on v0 are divergent'): # Call the base class constructor with the parameters it needs super(BoundError, self).__init__(message) class solver(): def __init__(self, gamma=1, a=10): """ Parameters: gamma: numerical constant, from equation of state a: numerical constant, equal to GM/cs^2 r0 """ self.gamma = gamma self.a = a self.W0 = np.log(1) def ndf1(self, psi, coords): '''Dimensionless f1 function, takes t, (x, y, z) and returns 1-exp[2y-z]''' try: #return 1-np.exp(2*coords[1]) return 1-np.exp(2*coords[1]-coords[2]) except OverflowError or UnderflowError: print("Error in ndf1 at ", psi, ", ", coords) return -1 def ndf2(self, psi, coords): '''Dimensionless f2 function, takes t, (x, y, z) and returns (GM/r0*cs^2)exp[-x-z]-2''' try: #return np.exp(-coords[0])*a-2 return self.a*np.exp(-coords[0]-coords[2])-2 except OverflowError or UnderflowError: print("Error in ndf2 at ", psi, ", ", coords) return -1 def dx(self, t, state): '''Infinitesimal x step''' return self.ndf1(t, state) #return state[0] def du(self, t, state): '''Infinitesimal u step''' return self.ndf2(t, state) #return state[1] def dw(self, t, state): '''Infinitesimal w step''' return -(self.gamma-1)*(2*self.ndf1(t, state)+self.ndf2(t, state)) #return state[2] def adx(self, t, state): '''Absolute value of dx''' return abs(self.ndf1(t, state)) def adu(self, t, state): '''Absolute value of du''' return abs(self.ndf2(t, state)) def adw(self, t, state): '''Absolute value of dw''' return -(self.gamma-1)*(2*abs(self.ndf1(t, state))+abs(self.ndf2(t, state))) def coupledRKStep(self, funs, state, dt): """RK4 step for array of coupled functions that evolve a state, for step size dt Returns array (t, new state) funs and state[1] should have the same length Parameters: funs: array of functions [dx, du, dw] to evolve a state state: numpy array [t, [x, u, w]] dt: generalized time step size Returns: numpy array [t, [x, u, w]] incremented by one RK4 step """ assert isinstance(funs, list), "Expected array input" for i in funs: assert callable(i), "Expected functions in list" assert isinstance(state, (np.ndarray, list)), "Expected array input" assert np.shape(state) == (2, ), "Expected 1D array with 2 elements" assert isinstance(state[1], (np.ndarray, list)), "Expected array in second element" assert np.shape(state[1]) == (len(funs), ), "State and function arrays are different lengths" assert isinstance(dt, (float, int, np.float64)), "Expected float input" karr = np.array([np.zeros(4) for i in range(len(funs))]) for j in range(4): for i in range(len(funs)): if j == 0: karr[i][j] = dt*funs[i](state[0], state[1]) else: karr[i][j] = dt*funs[i](state[0]+dt/2, state[1]+np.array([karr[k][j-1]/2 for k in range(len(karr))])) for i in range(len(funs)): karr[i][1] = 2*karr[i][1] karr[i][2] = 2*karr[i][2] step = np.array([state[1][i]+np.sum(karr[i])/6 for i in range(len(karr))]) return np.array([state[0]+dt, step]) def percentChange(self, curr, step): """Takes a current state and a new state (vectors) and calculates the percent change Parameters: curr: numpy array (x, u) step: numpy array (x', u') Returns: Percent change from curr to step""" assert isinstance(curr, (np.ndarray, list)), "Expected array input" assert isinstance(step, (np.ndarray, list)), "Expected array input" assert len(curr) == len(step), "Array lengths are not the same" return 100*abs(np.linalg.norm((step-curr)/np.linalg.norm(curr))) def adaptRK(self, currState, pc, t, dt, funs): """Iteratively adapts dt to ensure the change in (x, u) is between .1 and 1 percent This new dt can be either larger or smaller, independent of the initial percent change Takes a state (x, u, w), a percent change, t, dt, and the set of functions (dx, du, dw) being used to evolve the system Parameters: currState: numpy array (x, u, w) pc: percent change between currState and the next RK step with t, dt t: generalized time of currState dt: generalized time step size funs: array of functions [dx, du, dw] to evolve a state Returns: dt, adjusted so that the percent change is between .1 and 1 percent""" assert isinstance(funs, (np.ndarray, list)), "Expected array input" for i in funs: assert callable(i), "Expected functions in list" assert isinstance(currState, (np.ndarray, list)), "Expected array input" assert np.shape(currState) == (len(funs), ), "Expected 1D array input" assert isinstance(dt, (float, int, np.float64)), "Expected float input" assert isinstance(t, (float, int, np.float64)), "Expected float input, got {0}".format(type(t)) assert isinstance(pc, (float, int, np.float64)), "Expected float input" ddt = 1.5 i = 0 itermax = 10000 if pc > 1: pc2 = 1e10 #initialize dummy percent changes, used to track movement in % change while finding new dt prevpc2 = 1e10 while pc2 > 1 and i < itermax: dt = dt*ddt step2 = self.coupledRKStep(funs, [t, currState], dt) #calculate hypothetical next step using new dt pc2 = self.percentChange(currState, step2[1]) if pc2 > prevpc2: #if we're moving in the wrong direction, invert our change in dt ddt = 1/ddt prevpc2 = pc2 i = i+1 if i == itermax: print("Max iteration count exceeded in adaptation") return dt #once we've found a working dt, take a step using it elif pc < .1: pc2 = 1e-10 #initialize dummy percent changes, used to track movement in % change while finding new dt prevpc2 = 1e-10 while pc2 < .1 and i < itermax: dt = dt*ddt step2 = self.coupledRKStep(funs, [t, currState], dt) #calculate hypothetical next step using new dt pc2 = self.percentChange(currState, step2[1]) if pc2 < prevpc2: #if we're moving in the wrong direction, invert our change in dt ddt = 1/ddt prevpc2 = pc2 i = i+1 if i == itermax: print("Max iteration count exceeded in adaptation") return dt #once we've found a working dt, take a step using it def generateFunc(self, state0, itermax=10000, AV=True, xrange=10, urange=5): """Generates a trace of wind behavior with initial conditions x0 and u0 (dimensionless) using the RK4 method with adaptation in dt Takes x0, u0, max iteration count and returns a 2D array tracking t, x, and u Parameters: state0: initial state of system (x0, u0, w0) itermax: maximum iteration count for loop AV: use the absolute value of f1 and f2 (boolean) xrange: maximum x value to display on plot (displays (1, xrange)) urange: maximum u value to display on plot (displays (0, urange)) Returns: numpy array, [0] contains t values, [1] contains x values, [2] contains u values, [3] contains w values for the wind curve""" assert isinstance(state0, (list, np.ndarray)), "Expected array input" assert len(state0) == 3, "Expected state array of length 3" assert isinstance(itermax, int), "Expected integer input" assert itermax > 0, "Expected positive itermax" assert isinstance(AV, bool), "Expected boolean input" assert isinstance(xrange, (float, int, np.float64)), "Expected numerical input" assert xrange > 1, "Expected xrange > 1" assert isinstance(urange, (float, int, np.float64)), "Expected numerical input" assert urange > 0, "Expected positive urange" xsol = np.array([state0[0]]) usol = np.array([state0[1]]) wsol = np.array([state0[2]]) tarray = np.array([0]) t = 0 dt = .01 i = 0 currState = np.array([xsol[-1], usol[-1], wsol[-1]]) if AV: funs = [self.adx, self.adu, self.adw] else: funs = [self.dx, self.du, self.dw] #Main loop for adaptive RK solver #Exit conditions are based on values for exp(x) #Using zero points for f1 and f2 only works if you change ndf1 and ndf2 to return absolute values, and then you don't see the full curve #Setting a max iteration count doesn't always work well - the solution curves may be cut while np.exp(currState[0]) > 1e-6 and np.exp(currState[0]) < xrange and i < itermax: #Load the current position of the system to determine if adaptation is necessary currState = np.array([xsol[-1], usol[-1], wsol[-1]]) #Calculate the next integration step using the RK function defined above step = self.coupledRKStep(funs, [t, currState], dt) #Calculate percent change from current state to predicted next state pc = self.percentChange(currState, step[1]) #If the percent change is too large or too small, change dt to compensate if pc > 1 or pc < .1: dt = self.adaptRK(currState, pc, t, dt, funs) step = self.coupledRKStep(funs, [t, currState], dt) xsol = np.append(xsol, step[1][0]) #update solution curves with next step usol = np.append(usol, step[1][1]) wsol = np.append(wsol, step[1][2]) t = t+dt i = i+1 tarray = np.append(tarray, t) return np.array((tarray, xsol, usol, wsol)) def makePlot(self, u0, AV=True, xrange=10, urange=5): """Generates a plot of one wind curve Takes u0 and generates a curve No return, prints a plot of the wind curve Parameters: u0: initial value of u AV: use absolute value of f1 and f2 functions (boolean) xrange: maximum x value to display on plot (displays (1, xrange)) urange: maximum u value to display on plot (displays (0, urange)) Displays a plot of one wind curve""" assert isinstance(u0, (float, int, np.float64)), "Expected float input" assert u0 > 0, "Expected positive input" assert isinstance(AV, bool), "Expected boolean input" assert isinstance(xrange, (float, int, np.float64)), "Expected numerical input" assert xrange > 1, "Expected xrange > 1" assert isinstance(urange, (float, int, np.float64)), "Expected numerical input" assert urange > 0, "Expected positive urange" plt.figure(1) plt.xlim(1, xrange) plt.ylim(0, urange) func = self.generateFunc(np.array([0, np.log(u0), self.W0]), 10000, AV, xrange, urange) plt.scatter(np.exp(func[1]), np.exp(func[2]-func[3]/2), s=.5); plt.title("Velocity vs Radius (Dimensionless), v0 = %1.9f cs"%u0) plt.xlabel("r/r0") plt.ylabel("v/cs") def makePlots(self, vmin, vmax, dv, AV=True, xrange=10, urange=5, showAll=False): """Generates a plot of wind curves Takes vmin, vmax, dv and generates a curve for u0 (x0 = 0), then increments v0 by dv and generates a new curve, repeating until v0 = umax Expects vmin, vmax, and dv scaled by 1/cs No return, prints a plot of the different wind curves Parameters: umin: starting u value umax: maximum u value du: increment of u AV: Use absolute value of f1 and f2 functions (boolean) xrange: maximum x value to display on plot (displays (1, xrange)) urange: maximum u value to display on plot (displays (0, urange)) showAll: boolean, True will plot additional graphs of velocity and temperature Displays a plot of several different wind curves""" assert isinstance(vmin, (float, int, np.float64)), "Expected float input" assert vmin > 0, "Expected positive input" assert isinstance(vmax, (float, int, np.float64)), "Expected float input" assert vmax > 0, "Expected positive input" assert isinstance(dv, (float, int, np.float64)), "Expected float input" assert dv > 0, "Expected positive input" assert vmax > vmin, "Maximum less than minimum" assert isinstance(AV, bool), "Expected boolean input" assert isinstance(xrange, (float, int, np.float64)), "Expected numerical input" assert xrange > 1, "Expected xrange > 1" assert isinstance(urange, (float, int, np.float64)), "Expected numerical input" assert urange > 0, "Expected positive urange" plt.figure(1) plt.xlim(1, xrange) plt.ylim(0, urange) for i in np.arange(vmin, vmax, dv): func = self.generateFunc(np.array([0, np.log(i), self.W0]), 10000, AV, xrange, urange) if i == vmin: data = [func] else: data.append(func) plt.figure(1) plt.title("Velocity vs Radius (Dimensionless)") for i in range(len(data)): plt.scatter(np.exp(data[i][1]), np.exp(data[i][2]-data[i][3]/2), s=.5, label='v0/cs = %g'%np.exp(data[i][2][0])); plt.xlabel("r/r0") plt.ylabel("v/cs") plt.legend(loc="upper left", bbox_to_anchor=(1, 1)) if showAll: plt.figure(2) plt.title("Temperature vs Radius (Dimensionless)") for i in range(len(data)): plt.scatter(np.exp(data[i][1]), np.exp(data[i][3]), s = .5, label = 'v0/cs = %g'%np.exp(data[i][2][0])); plt.xlabel("r/r0") plt.ylabel("v/cs") plt.legend(loc = "upper left", bbox_to_anchor = (1, 1)) plt.figure(3) plt.title("Temperature vs Velocity (Dimensionless)") for i in range(len(data)): plt.scatter(np.exp(data[i][2]), np.exp(data[i][3]), s = .5, label = 'v0/cs = %g'%np.exp(data[i][2][0])); plt.xlabel("r/r0") plt.ylabel("v/cs") plt.legend(loc = "upper left", bbox_to_anchor = (1, 1)) def findZeros(self, v0): """Finds the t values where f1 and f2 reach zero, and returns the difference Takes starting velocity v0, expected to be scaled by 1/cs Returns tu-tx Parameters: v0: initial value of v/cs""" assert isinstance(v0, (float, int, np.float64)), "Expected numerical input" assert v0 > 0, "Expected positive input" u0 = np.log(v0) xsol = np.array([0]) usol = np.array([u0]) wsol = np.array([self.W0]) t = 0 dt = .01 currState = np.array([xsol[-1], usol[-1], wsol[-1]]) xfound = False ufound = False tu = 0 tx = 0 #Main loop, uses RK solver and iterates until f1 or f2 changes sign, and returns the t value where that takes place while not ufound and not xfound: #Load the current position of the system to determine if adaptation is necessary currState = np.array([xsol[-1], usol[-1], wsol[-1]]) #Calculate the next integration step using the RK function defined above step = self.coupledRKStep([self.adx, self.adu, self.adw], np.array([t, currState]), dt) #Calculate percent change from current state to predicted next state pc = self.percentChange(currState, step[1]) #If the percent change is too large or too small, change dt to compensate if pc > 1 or pc < .1: dt = self.adaptRK(currState, pc, t, dt, [self.adx, self.adu, self.adw]) step = self.coupledRKStep([self.adx, self.adu, self.adw], np.array([t, currState]), dt) #if ndf2 changes sign, we have found its turnover point and can exit the loop if np.sign(self.ndf2(t, step[1])) != np.sign(self.ndf2(t, currState)): xfound = True tx = t #if ndf1 changes sign, we have found its turnover point and can exit the loop if np.sign(self.ndf1(t, step[1])) != np.sign(self.ndf1(t, currState)): ufound = True tu = t #update solution curves with next step xsol = np.append(xsol, step[1][0]) usol = np.append(usol, step[1][1]) wsol = np.append(wsol, step[1][2]) t = t+dt #return difference between zeros in ndf1 and ndf2 return (tu-tx) def findVboundary(self, guess, increment=1e-4, maxprecision=1e-10, itermax=10000): """Locates the boundary value of v0 at which f1 and f2 pass through zero at the same time using a bisection method Parameters: guess: starting value of v0 increment: initial step size for incrementing v0 maxprecision: sets limit on how small dv can be before exiting the loop itermax: maximum iteration count Returns: Boundary value for v0""" assert isinstance(guess, (float, int, np.float64)), "Expected numerical input" assert guess > 0, "Expected positive input" assert isinstance(increment, (float, int, np.float64)), "Expected numerical input" assert isinstance(maxprecision, (float, int, np.float64)), "Expected numerical input" assert maxprecision > 0, "Expected positive input" assert isinstance(itermax, int), "Expected integer input" assert itermax > 0, "Expected positive input" v0 = guess dv = increment i = 0 #Main loop, while dv is larger than the max precision we want to continue refining our search while abs(dv) > maxprecision and i < itermax: #When we find a sign change in the zero for a given v0, that implies that the solution has changed character between a breeze and a non-physical one #i.e. whichever of f1 and f2 changed sign first, that has now reversed #when that happens, we shrink dv until we avoid the sign change, in order to approximate where exactly that occurs in v0 while (v0+dv) <= 0: dv = dv/2 self.findZeros(v0+dv) while np.sign(self.findZeros(v0+dv)) != np.sign(self.findZeros(v0)): dv = dv/2 while (v0+dv) <= 0: dv = dv/2 v0 = v0+dv i = i+1 if i >= itermax: print("Max iteration count exceeded") raise Exception("MaxIterExceeded") return v0+dv def findV0(self, lowerguess, upperguess, dv, maxprecision=1e-10, itermax=10000, xrange=10, urange=5, show=True, showPlot=False): """Finds boundary values for v0 and estimates an exact value for it Parameters: lowerguess: estimate of lower bound on v0/cs upperguess: estimate of upper bound on v0/cs dv: initial step size for incrementing v0/cs maxprecision: sets limit on how small dv can be before exiting the loop itermax: maximum iteration count xrange: maximum x value to display on plot (displays (1, xrange)) urange: maximum u value to display on plot (displays (0, urange)) show: show plot of v0 once it is found (boolean) Returns: Average of upper and lower bounds on v0 Prints bounds on v0, estimated value, and estimated error Displays plot of the wind curves for the boundary values of v0 if show = True""" assert isinstance(lowerguess, (float, int, np.float64)), "Expected numerical input" assert lowerguess > 0, "Expected positive input, got %"%lowerguess assert isinstance(upperguess, (float, int, np.float64)), "Expected numerical input" assert upperguess > 0, "Expected positive input, got %f"%upperguess assert isinstance(dv, (float, int, np.float64)), "Expected numerical input" assert dv > 0, "Expected positive input, got %"%dv assert isinstance(maxprecision, (float, int, np.float64)), "Expected numerical input" assert maxprecision > 0, "Expected positive input" assert isinstance(itermax, int), "Expected integer input" assert itermax > 0, "Expected positive input" assert isinstance(xrange, (float, int, np.float64)), "Expected numerical input" assert xrange > 1, "Expected xrange > 1" assert isinstance(urange, (float, int, np.float64)), "Expected numerical input" assert urange > 0, "Expected positive urange" if np.sign(self.findZeros(upperguess)) == np.sign(self.findZeros(lowerguess)): print("No sign change in specified range") raise NoSignChange dv=(upperguess-lowerguess)/2 upper = self.findVboundary(upperguess, -dv, maxprecision, itermax) lower = self.findVboundary(lowerguess, dv, maxprecision, itermax) if show: print("Lower bound on v0: ", lower) print("Upper bound on v0: ", upper) if abs(lower-upper) > 1e-2: print("divergent bounds, exiting") raise BoundError print("Estimated v0: ", (lower+upper)/2) print("Estimated error: ", abs((upper-lower)/2)) if showPlot: self.makePlot(lower, False, xrange, urange) self.makePlot(upper, False, xrange, urange) return (lower+upper)/2 def gammaSearch(self, a=10, g0=None, dg=.0025, glim=5/3, lower=1e-10, upper=.9, itermax=100): """Searches through gamma values for a given a and returns a table of gamma values and associated critical velocities Parameters: a: GM/cs^2 r0; constant value g0: starting gamma value, defaults to the gamma the class was initialized with dg: gamma increment (positive or negative) glim: gamma value at which the search will stop lower: estimate of lower bound on v0/cs for first gamma value upper: estimate of upper bound on v0/cs for first gamma value itermax: maximum iterations for searching a given gamma value for v0 Returns: List of ordered pairs [gamma, v0] Prints scatter plot of v0 vs gamma """ assert isinstance(a, (float, int, np.float64)), "Expected numerical input" assert a > 0, "Expected positive input" assert g0 is None or isinstance(g0, (float, int, np.float64)), "Expected numerical input" assert g0 is None or g0 > 0, "Expected positive input" assert isinstance(dg, (float, int, np.float64)), "Expected numerical input" assert isinstance(glim, (float, int, np.float64)), "Expected numerical input" assert isinstance(lower, (float, int, np.float64)), "Expected numerical input" assert lower > 0, "Expected positive input" assert isinstance(upper, (float, int, np.float64)), "Expected numerical input" assert upper > lower, "Nonsensical bounds" assert upper < 1, "Upper bound should be less than 1" gdata = np.array([]) if a != None: self.a = a if g0 != None: self.gamma = float(g0) else: g0 = float(self.gamma) i = 0 while (i < itermax) and ((self.gamma <= glim and np.sign(dg) == 1) or (self.gamma >= glim and np.sign(dg) == -1)): if i == 0: print("Searching gamma = ", self.gamma) if self.gamma == g0: try: with open(os.devnull, "w") as f, contextlib.redirect_stdout(f): gdata = [self.gamma, self.findV0(lower, upper, (upper-lower)/2, show = False)] self.gamma = self.gamma+dg i = 0 clear_output() except NoSignChange: if i == 0: print("No sign change, decrementing bounds") upper = float(lower) lower = lower/2 i = i+1 else: try: with open(os.devnull, "w") as f, contextlib.redirect_stdout(f): gdata = np.vstack((gdata, [self.gamma, self.findV0(lower, upper, (upper-lower)/2, show=False)])) self.gamma = self.gamma+dg i = 0 clear_output() except NoSignChange: if i == 0: print("No sign change, decrementing bounds") upper = float(lower) lower = lower/2 i = i+1 if i >= itermax: print("Max iteration count exceeded at gamma = ", self.gamma) print("No sign change above v0/cs = ", lower) if len(gdata) > 0: plt.figure(1) plt.title("Critical velocity vs. Gamma") plt.xlabel("Gamma") plt.ylabel("v0/cs") plt.scatter(gdata[:, 0], gdata[:, 1]) else: print("No data collected") return gdata
{ "alphanum_fraction": 0.5831949268, "author": null, "avg_line_length": 45.5189003436, "converted": null, "ext": "py", "file": null, "hexsha": "2cbe963f07942145345946724a3b259b4be4d736", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "0088a0568841cda00ee8303b797d05be9feab844", "max_forks_repo_licenses": [ "BSD-3-Clause" ], "max_forks_repo_name": "colbrydi/neutrino-winds", "max_forks_repo_path": "neutrino-winds/Adiabatic_wind_solver.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "0088a0568841cda00ee8303b797d05be9feab844", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "BSD-3-Clause" ], "max_issues_repo_name": "colbrydi/neutrino-winds", "max_issues_repo_path": "neutrino-winds/Adiabatic_wind_solver.py", "max_line_length": 160, "max_stars_count": null, "max_stars_repo_head_hexsha": "0088a0568841cda00ee8303b797d05be9feab844", "max_stars_repo_licenses": [ "BSD-3-Clause" ], "max_stars_repo_name": "colbrydi/neutrino-winds", "max_stars_repo_path": "neutrino-winds/Adiabatic_wind_solver.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 6644, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 26492 }
import abc import os import inspect import requests import pandas as pd import numpy as np import xarray as xr from .adapters import Adapter from .bounds import Bounds from .constants import (RCMED_QUERY_URL, ESGF_NODES, DEFAULT_INTAKE_CAT, DEFAULT_INTAKE_ESM_CAT) from .ensemble import Ensemble from .registry import registry, register from .utils import inherit_docs, decode_month_units, get_dropped_varnames class DataSource(object): """Loads datasets into ocw workspace. A data source describes where the input data comes from (eg, from your local filesystem or a remote server) and how to obtain it (via load()). All loaded datasets are represented in OCW as instances of xarray.DataArray. """ __metaclass__ = abc.ABCMeta def __init__(self, adapter=None): """Data source constructor. A data source describes where the input data comes from (eg, from your local filesystem or a remote server) and how to obtain it (via load()). All loaded datasets are represented in OCW as instances of xarray.DataArray. An optional parameter, adapter, allows for additional post-processing of the loaded DataArray object. The default adapter, an instance of bcdp.adapters.Adapter, simply changes the coordinate dimension names and attributes to comply with CF-format. For more complex post-processing use-cases, we recommend subclassing Adapter and passing an instance of such to this constructor. Parameters ---------- adapter : bcdp.Adapter Default adapter to use for post-processing. """ self.adapter = adapter if adapter else 'basic' self._cache = {} def __call__(self, *args, **kwargs): """Main interface for loading datasets. This combines the action of loading the dataset(s) from source into an `xarray.DataArray` object (via load()) and then building the evaluation Ensemble. Parameters ---------- adapter : str, optional Name of adapter class to use, which must be a valid key in the adapters registry. dims : dict, optional Mapping of dimension labels (x, y, z and/or t) to their respective names in the file(s). This should rarely ever be needed to be set, since these can be easily inferred most of the time. labels : dict, optional Mapping of labels **kwargs : dict Keyword arguments to pass to load(). Returns ------- bcdp.Ensemble List of datasets to be evaluated. """ dims = kwargs.pop('dims', {}) labels = kwargs.pop('labels', {}) datasets = self.load(*args, **kwargs) return Ensemble(datasets, adapter=self.adapter, dims=dims, **labels) @abc.abstractmethod def load(self, *args, **kwargs): """Loads datasets from given parameters.""" pass def _prep_datasets(self, variable, dset_dict): datasets = [] for name, ds in dset_dict.items(): variables = dict(ds.data_vars) if len(variables) == 1: # If no project or variable name information, infer it # from file_metadata. This will only work if the file has # one non-coordinate variable. variable = list(variables.keys())[0] elif not variable: raise ValueError('Variable name must be specified for files' ' with more than one non-coord variable.') # Check if variable name is defined in metadata. da = ds[variable].squeeze(drop=True) da.attrs['variable_name'] = da.name da.name = name datasets.append(da) return datasets @register('source.local') class LocalFileSource(DataSource): """Local Filesystem data source""" def load(self, paths='./*.nc', variable=None, names=None, convert_times=False, use_dt64=False, dims=None, project=None, load_all=False, **kwargs): """Loads datasets from given parameters. Parameters ---------- paths : str or list of str, optional Regex or (list of) full path(s) to the netcdf file(s) to open. variable : str, optional Variable Name. If input files have only one non-coordinate variable, that variable's name is used by default. names : list of str, optional List of dataset names. By default these are inferred directly from the input `paths` attribute. convert_times : bool, optional If True, files are assumed to be split by time and values are automatically converted to datetime-like objects. Does nothing if `project` is not set. use_dt64 : bool, optional If convert_times is True, use numpy.datetime64 to parse times instead of pandas.Timestamp. Only use this if the latter causes errors, as it is less flexible with datetimes that aren't in ISO 8601 format. project : str, optional A project name that encapsulates all of the datasets to be loaded. This must be a valid key in bcdp.extractors.metadata_extractors, which includes `CMIP5`, `CORDEX`, and `obs4MIPS`. When set, this replaces the default behavior for defining the variable and dataset names. For this reason, this parameter should only be set if you are sure that all of the input filenames correctly conform to the conventions required by the given project. load_all : bool, optional If True, datasets spanned by multiple files are loaded into memory rather than lazily (Default False). Ignored if chunks argument is passed to open_mfdataset. **kwargs Keyword Arguments to `xarray.open_(mf)dataset()` Returns ------- datasets : list xarray DataArray objects. """ if not isinstance(paths, list): paths = [paths] # Determine dataset names if not names: if project: # Use project specific conventions extractor_cls = registry['metadata'][project] extractor = extractor_cls(*paths) names = sorted(extractor.names) paths = sorted(extractor.files) else: # Infer dataset names from filenames pre = os.path.commonprefix(paths) post = os.path.commonprefix([p[::-1] for p in paths])[::-1] names = [p.replace(pre, '').replace(post, '') for p in paths] root_dir = os.path.dirname(pre) + os.path.sep paths = [root_dir + '*{}*'.format(name) for name in names] # Generic dataset loader def open_dataset(path, **kwargs): dropped = get_dropped_varnames(variable) if variable else None if os.path.isfile(path): ds = self._cache.get( path, xr.open_dataset(path, drop_variables=dropped, **kwargs) ) else: chunks = kwargs.pop('chunks', None) ds = self._cache.get( path, xr.open_mfdataset(path, drop_variables=dropped, **kwargs) ) ds = ds.chunk(chunks) if load_all and chunks is None: ds = ds.load() self._cache[path] = ds return ds # Open each dataset dset_dict = {} for name, path in zip(names, paths): try: ds = open_dataset(path, **kwargs) except ValueError: # Custom datetime decoding required for monthly time units kwargs.update(decode_times=False) ds = decode_month_units(open_dataset(path, **kwargs)) if project: # Get variable name from filename if project given. meta = extractor.query(filename=path)[0] if not variable: variable = meta['variable'] concat_dim = kwargs.pop('concat_dim', 'time') if concat_dim not in ds.coords: dim_vals = meta[concat_dim] if convert_times: if use_dt64: dim_vals = [np.datetime64(t) for t in dim_vals] else: dim_vals = [pd.Timestamp(t) for t in dim_vals] ds = ds.assign_coords({concat_dim: dim_vals}) dset_dict[name] = ds return self._prep_datasets(variable, dset_dict) @register('source.bucket') class BucketSource(DataSource): """"Load zarr data from cloud storage buckets.""" def load(self, *paths, variable=None, bucket_type=None, zarr_kwargs=None, **kwargs): """Loads datasets from given parameters. Parameters ---------- *paths : str Paths to zarr datastore in bucket. variable : str, optional Variable Name. If input files have only one non-coordinate variable, that variable's name is used by default. bucket_type : str, optional Type of cloud bucket (gc or s3). Can be ignored if the bucket type is in the url (eg gc://, s3://) zarr_kwargs : dict, optional Additional keyword arguments to pass to xarray.open_zarr. consolidated=True is set to default. **kwargs : dict, optional Keyword arguments to pass to bucke API (gcsfs or s3fs) Returns ------- datasets : list xarray DataArray objects. """ dset_dict = {} zarr_kwargs = zarr_kwargs if zarr_kwargs else dict(consolidated=True) for path in paths: if '://' in path: bucket_type = path.split('://')[0] if bucket_type not in ['s3', 'gc']: raise ValueError('Please specify supported bucket type (s3, gc)') if bucket_type == 's3': import s3fs fs = s3fs.S3FileSystem(**kwargs) mapper = s3fs.S3Map elif bucket_type == 'gc': import gcsfs fs = gcsfs.GCSFileSystem mapper = gcsfs.GCSMap if path.endswith('*'): sub_paths = fs.ls(path[:-2]) else: sub_paths = [path] for spath in sub_paths: name = spath.split('/')[-1] store = mapper(spath, fs) dset_dict[name] = xr.open_zarr(store, **zarr_kwargs) return self._prep_datasets(variable, dset_dict) @register('source.intake') class IntakeSource(DataSource): """"Load remote data via the intake library.""" def load(self, variable=None, names=None, depth=5, catfile=DEFAULT_INTAKE_CAT, auth_token=None): """Loads datasets from given parameters. Parameters ---------- variable : str, optional Variable Name. If input files have only one non-coordinate variable, that variable's name is used by default. names : list of str, optional List of dataset names. depth : int, optional Depth of catalog search (default: 5) catfile : str, optional Path to catalogue metadata file, can be a remote URL. The pangeo Intake master catalogue is used by default. auth_token : str, optional Path to credentials key file to use for accessing cloud storage buckets. Returns ------- datasets : list xarray DataArray objects. """ import intake if auth_token: os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = auth_token cat = intake.Catalog(catfile) meta = cat.walk(depth=depth) sel = [name for name, ent in meta.items() if ent.container == 'xarray'] names = sel if not names else names entries = [cat[name] for name in sel] shortnames = [name.split('.')[-1] for name in sel] dset_dict = {name: ent.to_dask() for name, ent in zip(shortnames, entries) if name in names} return self._prep_datasets(variable, dset_dict) @register('source.intake-esm') class IntakeESMSource(DataSource): """"Load remote data via the intake-esm library.""" def load(self, query, catfile=DEFAULT_INTAKE_ESM_CAT, **kwargs): """Loads datasets from given parameters. Parameters ---------- query: dict Key, value pairs used to search the catalogue. Depth of catalog search (default: 5) catfile : str, optional Path to catalogue metadata file, can be a remote URL. The pangeo intake-esm CMIP6 catalogue is used by default. **kwargs : dict, optional Keyword Arguments for `intake_esm.core.esm_datastore.to_dataset_dict()` Returns ------- datasets : list xarray DataArray objects. """ import intake col = intake.open_esm_datastore(catfile) cat = col.search(**query) dset_dict = cat.to_dataset_dict(**kwargs) variable = query.get('variable_id') return self._prep_datasets(variable, dset_dict) @register('source.rcmed') class RCMEDSource(DataSource): """JPL Regional Climate Model Evaluation Database (RCMED) Data Source""" def load(self, dataset_id, parameter_id, domain, chunks=None): """Loads datasets from given parameters. Parameters ---------- dataset_id : int Dataset ID in RCMED. variable_id : int Variable (Parameter) ID in RCMED. domain : bcdp.bounds.Bounds Bounds defining the spatial and temporal domain to search (and subset) the requested data. chunks : dict, optional If set, load data into a dask array with given chunk sizes. Returns ------- list of xarray.DataArray xarray DataArray object with loaded data. """ # Read RCMED metadata table from RCMES website metadata = self.get_metadata() idx = metadata.parameter_id == parameter_id if not idx.any(): raise ValueError(f'Invalid Parameter ID: {parameter_id}. ' 'Should be one of: {metadata.parameter_id}') # Format remaining parameters for request query params = dict(datasetId=dataset_id, parameterId=variable_id, **domain.to_dict()) # Make request to RCMED server r = requests.get(RCMED_QUERY_URL, params=params) r.raise_for_status() result = r.text.split('\r\n') meta = result[:12] data = result[12:-1] # Create DataArray object from output variable_name = meta[2].split('Parameter:')[1].split('(')[0].strip() name = metadata.database[idx].values[0] df = pd.DataFrame(data=data) df = df[0].str.split(',', expand=True) numeric_dims = ['lat', 'lon', 'lev', variable_name] df.columns = ['lat', 'lon', 'lev', 'time', variable_name] df[numeric_dims] = df[numeric_dims].astype(float) df['time'] = pd.to_datetime(df['time']) df = df.set_index(['time', 'lev', 'lat', 'lon']) da = df[variable_name].to_xarray().squeeze() if chunks: da = da.chunk(chunks) da.attrs['units'] = metadata['units'][idx].values[0] da = da.where(da != metadata['missingdataflag'][idx].values[0]) da.name = name return [da] def get_metadata(self): """Loads RCMED metadata. Returns ------- meta : pandas.DataFrame RCMED metadata table. """ info = requests.get(RCMED_QUERY_URL, params=dict(param_info='yes')).json() meta = pd.DataFrame(data=info['data'], columns=info['fields_name']) return meta @register('source.esgf') class ESGFSource(DataSource): """Earth System Grid (ESGF) data source""" _origin = 'esgf' def load(self, variable=None, project=None, node='JPL', **kwargs): """Builds an xarray.DataArray object from given parameters. Parameters ---------- paths : str or list of str, optional Regex or (list of) full path(s) to the netcdf file(s) to open. variable : str, optional Variable Name. project : str, optional A project name that encapsulates all of the datasets to be loaded. This must be a valid key in the bcdp object registry, which includes `CMIP5`, `CORDEX`, and `obs4MIPS`. When set, this replaces the default behavior for defining the variable and dataset names. For this reason, this parameter should only be set if you are sure that all of the input filenames correctly conform to the conventions required by the given project. node : str, optional ESGF node to search from, default is JPL. **kwargs : dict, optional ESGF search parameters. Returns ------- datasets : dict of xarray.DataArray xarray DataArray objects with origin information and CF-compliant coordinate variables, keyed by dataset names. """ hostname = ESGF_NODES[node] raise NotImplementedError('') load_local = LocalFileSource() load_bucket = BucketSource() load_intake = IntakeSource() load_intake_esm = IntakeESMSource() load_rcmed = RCMEDSource()
{ "alphanum_fraction": 0.5699197068, "author": null, "avg_line_length": 40.8744493392, "converted": null, "ext": "py", "file": null, "hexsha": "317bd89efcfdb25fcfe273e43c8ee7ba9bfec5c9", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 2, "max_forks_repo_forks_event_max_datetime": "2020-04-04T09:33:00.000Z", "max_forks_repo_forks_event_min_datetime": "2020-02-05T23:28:32.000Z", "max_forks_repo_head_hexsha": "63e8b93251599cbb591567459d7b598472c6afc8", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "bcdp/bcdp", "max_forks_repo_path": "bcdp/sources.py", "max_issues_count": 1, "max_issues_repo_head_hexsha": "63e8b93251599cbb591567459d7b598472c6afc8", "max_issues_repo_issues_event_max_datetime": "2020-04-16T22:17:45.000Z", "max_issues_repo_issues_event_min_datetime": "2020-04-16T22:17:45.000Z", "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "bcdp/bcdp", "max_issues_repo_path": "bcdp/sources.py", "max_line_length": 84, "max_stars_count": 5, "max_stars_repo_head_hexsha": "63e8b93251599cbb591567459d7b598472c6afc8", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "bcdp/bcdp", "max_stars_repo_path": "bcdp/sources.py", "max_stars_repo_stars_event_max_datetime": "2021-09-16T14:58:00.000Z", "max_stars_repo_stars_event_min_datetime": "2020-02-17T10:24:32.000Z", "num_tokens": 3835, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 18557 }
[STATEMENT] lemma numsubst0_numbound0: assumes "numbound0 t" shows "numbound0 (numsubst0 t a)" [PROOF STATE] proof (prove) goal (1 subgoal): 1. numbound0 (numsubst0 t a) [PROOF STEP] using assms [PROOF STATE] proof (prove) using this: numbound0 t goal (1 subgoal): 1. numbound0 (numsubst0 t a) [PROOF STEP] proof (induct a) [PROOF STATE] proof (state) goal (7 subgoals): 1. \<And>x. numbound0 t \<Longrightarrow> numbound0 (numsubst0 t (C x)) 2. \<And>x. numbound0 t \<Longrightarrow> numbound0 (numsubst0 t (Bound x)) 3. \<And>x1a x2a a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (CN x1a x2a a)) 4. \<And>a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Neg a)) 5. \<And>a1 a2. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a1); numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a2); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Add a1 a2)) 6. \<And>a1 a2. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a1); numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a2); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Sub a1 a2)) 7. \<And>x1a a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Mul x1a a)) [PROOF STEP] case (CN n) [PROOF STATE] proof (state) this: numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a_) numbound0 t goal (7 subgoals): 1. \<And>x. numbound0 t \<Longrightarrow> numbound0 (numsubst0 t (C x)) 2. \<And>x. numbound0 t \<Longrightarrow> numbound0 (numsubst0 t (Bound x)) 3. \<And>x1a x2a a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (CN x1a x2a a)) 4. \<And>a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Neg a)) 5. \<And>a1 a2. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a1); numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a2); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Add a1 a2)) 6. \<And>a1 a2. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a1); numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a2); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Sub a1 a2)) 7. \<And>x1a a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Mul x1a a)) [PROOF STEP] then [PROOF STATE] proof (chain) picking this: numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a_) numbound0 t [PROOF STEP] show ?case [PROOF STATE] proof (prove) using this: numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a_) numbound0 t goal (1 subgoal): 1. numbound0 (numsubst0 t (CN n x2a_ a_)) [PROOF STEP] by (cases n) simp_all [PROOF STATE] proof (state) this: numbound0 (numsubst0 t (CN n x2a_ a_)) goal (6 subgoals): 1. \<And>x. numbound0 t \<Longrightarrow> numbound0 (numsubst0 t (C x)) 2. \<And>x. numbound0 t \<Longrightarrow> numbound0 (numsubst0 t (Bound x)) 3. \<And>a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Neg a)) 4. \<And>a1 a2. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a1); numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a2); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Add a1 a2)) 5. \<And>a1 a2. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a1); numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a2); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Sub a1 a2)) 6. \<And>x1a a. \<lbrakk>numbound0 t \<Longrightarrow> numbound0 (numsubst0 t a); numbound0 t\<rbrakk> \<Longrightarrow> numbound0 (numsubst0 t (Mul x1a a)) [PROOF STEP] qed simp_all
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": null, "hexsha": null, "include": null, "lang": null, "length": 7, "llama_tokens": 1565, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
function vertCongruenceAA(W) err, i, todelete, vclasses = 10^-6, 1, [], [] verts = convert(Lar.Points, W') kdtree = NearestNeighbors.KDTree(verts) newverts = zeros(Int, size(verts,2)) for vi in 1:size(verts,2) if !(vi in todelete) nearvs = NearestNeighbors.inrange(kdtree, verts[:,vi], err) push!(vclasses,nearvs) newverts[nearvs] .= i nearvs = setdiff(nearvs, vi) todelete = union(todelete, nearvs) i += 1 end end V = zeros(3,length(vclasses)) for (k,class) in enumerate(vclasses) V[:,k] = sum(W[class,:],dims=1)/length(class) end return V, vclasses end function cellCongruenceAA(Delta,inclasses) cellarray = Lar.cop2lar(Delta) new_e = Array{Int64,1}(undef,size(Delta,2)) for (k,class) in enumerate(inclasses) for e in class new_e[e] = k end end cells = [map(e->new_e[e], face) for face in cellarray] outclasses = DefaultOrderedDict{Array{Int64,1},Array{Int64,1}}([]) for (k,face) in enumerate(cells) if outclasses[face] == [] outclasses[face] = [k] else append!(outclasses[face],[k]) end end FEnew = sort(collect(keys(outclasses))) outclasses = sort(collect(values(outclasses))) return FEnew,outclasses end function chainCongruenceAA(W, T) V, vclasses = vertCongruenceAA(W) EV, eclasses = cellCongruenceAA(T[1],vclasses) FE, fclasses = cellCongruenceAA(T[2],eclasses) #@show size(V,2) - size(EV,1) + size(FE,1) return V,EV,FE end
{ "alphanum_fraction": 0.6793248945, "author": null, "avg_line_length": 26.3333333333, "converted": null, "ext": "jl", "file": null, "hexsha": "c37df422ab2d06863d2f38341123e15a0daffa3a", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2022-03-31T11:20:41.000Z", "max_forks_repo_forks_event_min_datetime": "2022-03-31T11:20:41.000Z", "max_forks_repo_head_hexsha": "3b8d254681d56fab1fbdf305a36dd4e3234c7119", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "cvdlab/LocalCongruence.jl", "max_forks_repo_path": "src/cea-AA.jl", "max_issues_count": null, "max_issues_repo_head_hexsha": "3b8d254681d56fab1fbdf305a36dd4e3234c7119", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "cvdlab/LocalCongruence.jl", "max_issues_repo_path": "src/cea-AA.jl", "max_line_length": 68, "max_stars_count": null, "max_stars_repo_head_hexsha": "3b8d254681d56fab1fbdf305a36dd4e3234c7119", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "cvdlab/LocalCongruence.jl", "max_stars_repo_path": "src/cea-AA.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 484, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 1422 }
const INTERNALNAMES = (:__model__, :__context__, :__varinfo__) """ isassumption(expr[, vn]) Return an expression that can be evaluated to check if `expr` is an assumption in the model. Let `expr` be `:(x[1])`. It is an assumption in the following cases: 1. `x` is not among the input data to the model, 2. `x` is among the input data to the model but with a value `missing`, or 3. `x` is among the input data to the model with a value other than missing, but `x[1] === missing`. When `expr` is not an expression or symbol (i.e., a literal), this expands to `false`. If `vn` is specified, it will be assumed to refer to a expression which evaluates to a `VarName`, and this will be used in the subsequent checks. If `vn` is not specified, `AbstractPPL.drop_escape(varname(expr))` will be used in its place. """ function isassumption(expr::Union{Expr,Symbol}, vn=AbstractPPL.drop_escape(varname(expr))) return quote if $(DynamicPPL.contextual_isassumption)(__context__, $vn) # Considered an assumption by `__context__` which means either: # 1. We hit the default implementation, e.g. using `DefaultContext`, # which in turn means that we haven't considered if it's one of # the model arguments, hence we need to check this. # 2. We are working with a `ConditionContext` _and_ it's NOT in the model arguments, # i.e. we're trying to condition one of the latent variables. # In this case, the below will return `true` since the first branch # will be hit. # 3. We are working with a `ConditionContext` _and_ it's in the model arguments, # i.e. we're trying to override the value. This is currently NOT supported. # TODO: Support by adding context to model, and use `model.args` # as the default conditioning. Then we no longer need to check `inargnames` # since it will all be handled by `contextual_isassumption`. if !($(DynamicPPL.inargnames)($vn, __model__)) || $(DynamicPPL.inmissings)($vn, __model__) true else $(maybe_view(expr)) === missing end else false end end end # failsafe: a literal is never an assumption isassumption(expr, vn) = :(false) isassumption(expr) = :(false) """ contextual_isassumption(context, vn) Return `true` if `vn` is considered an assumption by `context`. The default implementation for `AbstractContext` always returns `true`. """ contextual_isassumption(::IsLeaf, context, vn) = true function contextual_isassumption(::IsParent, context, vn) return contextual_isassumption(childcontext(context), vn) end function contextual_isassumption(context::AbstractContext, vn) return contextual_isassumption(NodeTrait(context), context, vn) end function contextual_isassumption(context::ConditionContext, vn) if hasvalue(context, vn) val = getvalue(context, vn) # TODO: Do we even need the `>: Missing`, i.e. does it even help the compiler? if eltype(val) >: Missing && val === missing return true else return false end end # We might have nested contexts, e.g. `ContextionContext{.., <:PrefixContext{..., <:ConditionContext}}` # so we defer to `childcontext` if we haven't concluded that anything yet. return contextual_isassumption(childcontext(context), vn) end function contextual_isassumption(context::PrefixContext, vn) return contextual_isassumption(childcontext(context), prefix(context, vn)) end # If we're working with, say, a `Symbol`, then we're not going to `view`. maybe_view(x) = x maybe_view(x::Expr) = :(@views($x)) """ isliteral(expr) Return `true` if `expr` is a literal, e.g. `1.0` or `[1.0, ]`, and `false` otherwise. """ isliteral(e) = false isliteral(::Number) = true isliteral(e::Expr) = !isempty(e.args) && all(isliteral, e.args) """ check_tilde_rhs(x) Check if the right-hand side `x` of a `~` is a `Distribution` or an array of `Distributions`, then return `x`. """ function check_tilde_rhs(@nospecialize(x)) return throw( ArgumentError( "the right-hand side of a `~` must be a `Distribution` or an array of `Distribution`s", ), ) end check_tilde_rhs(x::Distribution) = x check_tilde_rhs(x::AbstractArray{<:Distribution}) = x """ unwrap_right_vn(right, vn) Return the unwrapped distribution on the right-hand side and variable name on the left-hand side of a `~` expression such as `x ~ Normal()`. This is used mainly to unwrap `NamedDist` distributions. """ unwrap_right_vn(right, vn) = right, vn unwrap_right_vn(right::NamedDist, vn) = unwrap_right_vn(right.dist, right.name) """ unwrap_right_left_vns(right, left, vns) Return the unwrapped distributions on the right-hand side and values and variable names on the left-hand side of a `.~` expression such as `x .~ Normal()`. This is used mainly to unwrap `NamedDist` distributions and adjust the indices of the variables. # Example ```jldoctest; setup=:(using Distributions, LinearAlgebra) julia> _, _, vns = DynamicPPL.unwrap_right_left_vns(MvNormal(ones(2), I), randn(2, 2), @varname(x)); vns[end] x[:,2] julia> _, _, vns = DynamicPPL.unwrap_right_left_vns(Normal(), randn(1, 2), @varname(x)); vns[end] x[1,2] julia> _, _, vns = DynamicPPL.unwrap_right_left_vns(Normal(), randn(1, 2), @varname(x[:])); vns[end] x[:][1,2] julia> _, _, vns = DynamicPPL.unwrap_right_left_vns(Normal(), randn(3), @varname(x[1])); vns[end] x[1][3] ``` """ unwrap_right_left_vns(right, left, vns) = right, left, vns function unwrap_right_left_vns(right::NamedDist, left, vns) return unwrap_right_left_vns(right.dist, left, right.name) end function unwrap_right_left_vns( right::MultivariateDistribution, left::AbstractMatrix, vn::VarName ) # This an expression such as `x .~ MvNormal()` which we interpret as # x[:, i] ~ MvNormal() # for `i = size(left, 2)`. Hence the symbol should be `x[:, i]`, # and we therefore add the `Colon()` below. vns = map(axes(left, 2)) do i return vn ∘ Setfield.IndexLens((Colon(), i)) end return unwrap_right_left_vns(right, left, vns) end function unwrap_right_left_vns( right::Union{Distribution,AbstractArray{<:Distribution}}, left::AbstractArray, vn::VarName, ) vns = map(CartesianIndices(left)) do i return vn ∘ Setfield.IndexLens(Tuple(i)) end return unwrap_right_left_vns(right, left, vns) end ################# # Main Compiler # ################# """ @model(expr[, warn = false]) Macro to specify a probabilistic model. If `warn` is `true`, a warning is displayed if internal variable names are used in the model definition. # Examples Model definition: ```julia @model function model(x, y = 42) ... end ``` To generate a `Model`, call `model(xvalue)` or `model(xvalue, yvalue)`. """ macro model(expr, warn=false) # include `LineNumberNode` with information about the call site in the # generated function for easier debugging and interpretation of error messages return esc(model(__module__, __source__, expr, warn)) end function model(mod, linenumbernode, expr, warn) modelinfo = build_model_info(expr) # Generate main body modelinfo[:body] = generate_mainbody(mod, modelinfo[:modeldef][:body], warn) return build_output(modelinfo, linenumbernode) end """ build_model_info(input_expr) Builds the `model_info` dictionary from the model's expression. """ function build_model_info(input_expr) # Break up the model definition and extract its name, arguments, and function body modeldef = MacroTools.splitdef(input_expr) # Check that the function has a name # https://github.com/TuringLang/DynamicPPL.jl/issues/260 haskey(modeldef, :name) || throw(ArgumentError("anonymous functions without name are not supported")) # Print a warning if function body of the model is empty warn_empty(modeldef[:body]) ## Construct model_info dictionary # Shortcut if the model does not have any arguments if !haskey(modeldef, :args) && !haskey(modeldef, :kwargs) modelinfo = Dict( :allargs_exprs => [], :allargs_syms => [], :allargs_namedtuple => NamedTuple(), :defaults_namedtuple => NamedTuple(), :modeldef => modeldef, ) return modelinfo end # Ensure that all arguments have a name, i.e., are of the form `name` or `name::T` addargnames!(modeldef[:args]) # Extract the positional and keyword arguments from the model definition. allargs = vcat(modeldef[:args], modeldef[:kwargs]) # Split the argument expressions and the default values. allargs_exprs_defaults = map(allargs) do arg MacroTools.@match arg begin (x_ = val_) => (x, val) x_ => (x, NO_DEFAULT) end end # Extract the expressions of the arguments, without default values. allargs_exprs = first.(allargs_exprs_defaults) # Extract the names of the arguments. allargs_syms = map(allargs_exprs) do arg MacroTools.@match arg begin (name_::_) => name x_ => x end end # Build named tuple expression of the argument symbols and variables of the same name. allargs_namedtuple = to_namedtuple_expr(allargs_syms) # Extract default values of the positional and keyword arguments. default_syms = [] default_vals = [] for (sym, (expr, val)) in zip(allargs_syms, allargs_exprs_defaults) if val !== NO_DEFAULT push!(default_syms, sym) push!(default_vals, val) end end # Build named tuple expression of the argument symbols with default values. defaults_namedtuple = to_namedtuple_expr(default_syms, default_vals) modelinfo = Dict( :allargs_exprs => allargs_exprs, :allargs_syms => allargs_syms, :allargs_namedtuple => allargs_namedtuple, :defaults_namedtuple => defaults_namedtuple, :modeldef => modeldef, ) return modelinfo end """ generate_mainbody(mod, expr, warn) Generate the body of the main evaluation function from expression `expr` and arguments `args`. If `warn` is true, a warning is displayed if internal variables are used in the model definition. """ generate_mainbody(mod, expr, warn) = generate_mainbody!(mod, Symbol[], expr, warn) generate_mainbody!(mod, found, x, warn) = x function generate_mainbody!(mod, found, sym::Symbol, warn) if warn && sym in INTERNALNAMES && sym ∉ found @warn "you are using the internal variable `$sym`" push!(found, sym) end return sym end function generate_mainbody!(mod, found, expr::Expr, warn) # Do not touch interpolated expressions expr.head === :$ && return expr.args[1] # Do we don't want escaped expressions because we unfortunately # escape the entire body afterwards. Meta.isexpr(expr, :escape) && return generate_mainbody(mod, found, expr.args[1], warn) # If it's a macro, we expand it if Meta.isexpr(expr, :macrocall) return generate_mainbody!(mod, found, macroexpand(mod, expr; recursive=true), warn) end # Modify dotted tilde operators. args_dottilde = getargs_dottilde(expr) if args_dottilde !== nothing L, R = args_dottilde return Base.remove_linenums!( generate_dot_tilde( generate_mainbody!(mod, found, L, warn), generate_mainbody!(mod, found, R, warn), ), ) end # Modify tilde operators. args_tilde = getargs_tilde(expr) if args_tilde !== nothing L, R = args_tilde return Base.remove_linenums!( generate_tilde( generate_mainbody!(mod, found, L, warn), generate_mainbody!(mod, found, R, warn), ), ) end return Expr(expr.head, map(x -> generate_mainbody!(mod, found, x, warn), expr.args)...) end function generate_tilde_literal(left, right) # If the LHS is a literal, it is always an observation @gensym value return quote $value, __varinfo__ = $(DynamicPPL.tilde_observe!!)( __context__, $(DynamicPPL.check_tilde_rhs)($right), $left, __varinfo__ ) $value end end """ generate_tilde(left, right) Generate an `observe` expression for data variables and `assume` expression for parameter variables. """ function generate_tilde(left, right) isliteral(left) && return generate_tilde_literal(left, right) # Otherwise it is determined by the model or its value, # if the LHS represents an observation @gensym vn isassumption value # HACK: Usage of `drop_escape` is unfortunate. It's a consequence of the fact # that in DynamicPPL we the entire function body. Instead we should be # more selective with our escape. Until that's the case, we remove them all. return quote $vn = $(AbstractPPL.drop_escape(varname(left))) $isassumption = $(DynamicPPL.isassumption(left, vn)) if $isassumption $(generate_tilde_assume(left, right, vn)) else # If `vn` is not in `argnames`, we need to make sure that the variable is defined. if !$(DynamicPPL.inargnames)($vn, __model__) $left = $(DynamicPPL.getvalue_nested)(__context__, $vn) end $value, __varinfo__ = $(DynamicPPL.tilde_observe!!)( __context__, $(DynamicPPL.check_tilde_rhs)($right), $(maybe_view(left)), $vn, __varinfo__, ) $value end end end function generate_tilde_assume(left, right, vn) # HACK: Because the Setfield.jl macro does not support assignment # with multiple arguments on the LHS, we need to capture the return-values # and then update the LHS variables one by one. @gensym value expr = :($left = $value) if left isa Expr expr = AbstractPPL.drop_escape( Setfield.setmacro(BangBang.prefermutation, expr; overwrite=true) ) end return quote $value, __varinfo__ = $(DynamicPPL.tilde_assume!!)( __context__, $(DynamicPPL.unwrap_right_vn)($(DynamicPPL.check_tilde_rhs)($right), $vn)..., __varinfo__, ) $expr $value end end """ generate_dot_tilde(left, right) Generate the expression that replaces `left .~ right` in the model body. """ function generate_dot_tilde(left, right) isliteral(left) && return generate_tilde_literal(left, right) # Otherwise it is determined by the model or its value, # if the LHS represents an observation @gensym vn isassumption value return quote $vn = $(AbstractPPL.drop_escape(varname(left))) $isassumption = $(DynamicPPL.isassumption(left, vn)) if $isassumption $(generate_dot_tilde_assume(left, right, vn)) else # If `vn` is not in `argnames`, we need to make sure that the variable is defined. if !$(DynamicPPL.inargnames)($vn, __model__) $left .= $(DynamicPPL.getvalue_nested)(__context__, $vn) end $value, __varinfo__ = $(DynamicPPL.dot_tilde_observe!!)( __context__, $(DynamicPPL.check_tilde_rhs)($right), $(maybe_view(left)), $vn, __varinfo__, ) $value end end end function generate_dot_tilde_assume(left, right, vn) # We don't need to use `Setfield.@set` here since # `.=` is always going to be inplace + needs `left` to # be something that supports `.=`. @gensym value return quote $value, __varinfo__ = $(DynamicPPL.dot_tilde_assume!!)( __context__, $(DynamicPPL.unwrap_right_left_vns)( $(DynamicPPL.check_tilde_rhs)($right), $(maybe_view(left)), $vn )..., __varinfo__, ) $left .= $value $value end end # Note that we cannot use `MacroTools.isdef` because # of https://github.com/FluxML/MacroTools.jl/issues/154. """ isfuncdef(expr) Return `true` if `expr` is any form of function definition, and `false` otherwise. """ function isfuncdef(e::Expr) return if Meta.isexpr(e, :function) # Classic `function f(...)` true elseif Meta.isexpr(e, :->) # Anonymous functions/lambdas, e.g. `do` blocks or `->` defs. true elseif Meta.isexpr(e, :(=)) && Meta.isexpr(e.args[1], :call) # Short function defs, e.g. `f(args...) = ...`. true else false end end """ replace_returns(expr) Return `Expr` with all `return ...` statements replaced with `return ..., DynamicPPL.return_values(__varinfo__)`. Note that this method will _not_ replace `return` statements within function definitions. This is checked using [`isfuncdef`](@ref). """ replace_returns(e) = e function replace_returns(e::Expr) if isfuncdef(e) return e end if Meta.isexpr(e, :return) # We capture the original return-value in `retval` and return # a `Tuple{typeof(retval),typeof(__varinfo__)}`. # If we don't capture the return-value separately, cases such as # `return x = 1` will result in `(x = 1, __varinfo__)` which will # mistakenly attempt to construct a `NamedTuple` (which fails on Julia 1.3 # and is not our intent). @gensym retval return quote $retval = $(e.args...) return $retval, __varinfo__ end end return Expr(e.head, map(replace_returns, e.args)...) end # If it's just a symbol, e.g. `f(x) = 1`, then we make it `f(x) = return 1`. make_returns_explicit!(body) = Expr(:return, body) function make_returns_explicit!(body::Expr) # If the last statement is a return-statement, we don't do anything. # Otherwise we replace the last statement with a `return` statement. if !Meta.isexpr(body.args[end], :return) body.args[end] = Expr(:return, body.args[end]) end return body end const FloatOrArrayType = Type{<:Union{AbstractFloat,AbstractArray}} hasmissing(::Type) = false hasmissing(::Type{>:Missing}) = true hasmissing(::Type{<:AbstractArray{TA}}) where {TA} = hasmissing(TA) hasmissing(::Type{Union{}}) = false # issue #368 """ build_output(modelinfo, linenumbernode) Builds the output expression. """ function build_output(modelinfo, linenumbernode) ## Build the anonymous evaluator from the user-provided model definition. evaluatordef = deepcopy(modelinfo[:modeldef]) # Add the internal arguments to the user-specified arguments (positional + keywords). evaluatordef[:args] = vcat( [ :(__model__::$(DynamicPPL.Model)), :(__varinfo__::$(DynamicPPL.AbstractVarInfo)), :(__context__::$(DynamicPPL.AbstractContext)), ], modelinfo[:allargs_exprs], ) # Delete the keyword arguments. evaluatordef[:kwargs] = [] # Replace the user-provided function body with the version created by DynamicPPL. # We use `MacroTools.@q begin ... end` instead of regular `quote ... end` to ensure # that no new `LineNumberNode`s are added apart from the reference `linenumbernode` # to the call site. # NOTE: We need to replace statements of the form `return ...` with # `return (..., __varinfo__)` to ensure that the second # element in the returned value is always the most up-to-date `__varinfo__`. # See the docstrings of `replace_returns` for more info. evaluatordef[:body] = MacroTools.@q begin $(linenumbernode) $(replace_returns(make_returns_explicit!(modelinfo[:body]))) end ## Build the model function. # Extract the named tuple expression of all arguments and the default values. allargs_namedtuple = modelinfo[:allargs_namedtuple] defaults_namedtuple = modelinfo[:defaults_namedtuple] # Obtain or generate the name of the model to support functors: # https://github.com/TuringLang/DynamicPPL.jl/issues/367 modeldef = modelinfo[:modeldef] if MacroTools.@capture(modeldef[:name], ::T_) name = gensym(:f) modeldef[:name] = Expr(:(::), name, T) elseif MacroTools.@capture(modeldef[:name], (name_::_ | name_)) else throw(ArgumentError("unsupported format of model function")) end # Update the function body of the user-specified model. # We use `MacroTools.@q begin ... end` instead of regular `quote ... end` to ensure # that no new `LineNumberNode`s are added apart from the reference `linenumbernode` # to the call site modeldef[:body] = MacroTools.@q begin $(linenumbernode) return $(DynamicPPL.Model)($name, $allargs_namedtuple, $defaults_namedtuple) end return MacroTools.@q begin $(MacroTools.combinedef(evaluatordef)) $(Base).@__doc__ $(MacroTools.combinedef(modeldef)) end end function warn_empty(body) if all(l -> isa(l, LineNumberNode), body.args) @warn("Model definition seems empty, still continue.") end return nothing end """ matchingvalue(sampler, vi, value) matchingvalue(context::AbstractContext, vi, value) Convert the `value` to the correct type for the `sampler` or `context` and the `vi` object. For a `context` that is _not_ a `SamplingContext`, we fall back to `matchingvalue(SampleFromPrior(), vi, value)`. """ function matchingvalue(sampler, vi, value) T = typeof(value) if hasmissing(T) _value = convert(get_matching_type(sampler, vi, T), value) if _value === value return deepcopy(_value) else return _value end else return value end end function matchingvalue(sampler::AbstractSampler, vi, value::FloatOrArrayType) return get_matching_type(sampler, vi, value) end function matchingvalue(context::AbstractContext, vi, value) return matchingvalue(NodeTrait(matchingvalue, context), context, vi, value) end function matchingvalue(::IsLeaf, context::AbstractContext, vi, value) return matchingvalue(SampleFromPrior(), vi, value) end function matchingvalue(::IsParent, context::AbstractContext, vi, value) return matchingvalue(childcontext(context), vi, value) end function matchingvalue(context::SamplingContext, vi, value) return matchingvalue(context.sampler, vi, value) end """ get_matching_type(spl::AbstractSampler, vi, ::Type{T}) where {T} Get the specialized version of type `T` for sampler `spl`. For example, if `T === Float64` and `spl::Hamiltonian`, the matching type is `eltype(vi[spl])`. """ get_matching_type(spl::AbstractSampler, vi, ::Type{T}) where {T} = T function get_matching_type(spl::AbstractSampler, vi, ::Type{<:Union{Missing,AbstractFloat}}) return Union{Missing,floatof(eltype(vi, spl))} end function get_matching_type(spl::AbstractSampler, vi, ::Type{<:AbstractFloat}) return floatof(eltype(vi, spl)) end function get_matching_type(spl::AbstractSampler, vi, ::Type{<:Array{T,N}}) where {T,N} return Array{get_matching_type(spl, vi, T),N} end function get_matching_type(spl::AbstractSampler, vi, ::Type{<:Array{T}}) where {T} return Array{get_matching_type(spl, vi, T)} end floatof(::Type{T}) where {T<:Real} = typeof(one(T) / one(T)) floatof(::Type) = Real # fallback if type inference failed
{ "alphanum_fraction": 0.6598417869, "author": null, "avg_line_length": 34.1111111111, "converted": null, "ext": "jl", "file": null, "hexsha": "4038d5d14fa21e2ffffb7b218e417419814bf00a", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "d222316a7a2fd5afe6ec74a7ec2a50c6f08c1d00", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "TuringLang/MicroPPL", "max_forks_repo_path": "src/compiler.jl", "max_issues_count": null, "max_issues_repo_head_hexsha": "d222316a7a2fd5afe6ec74a7ec2a50c6f08c1d00", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "TuringLang/MicroPPL", "max_issues_repo_path": "src/compiler.jl", "max_line_length": 109, "max_stars_count": null, "max_stars_repo_head_hexsha": "d222316a7a2fd5afe6ec74a7ec2a50c6f08c1d00", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "TuringLang/MicroPPL", "max_stars_repo_path": "src/compiler.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 6020, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 23639 }
import os, sys lib_path = os.path.abspath(os.path.dirname(__file__)) sys.path.append(lib_path) import tensorflow as tf from ddpg_actor import DDPG_Actor from ddpg_critic import DDPG_Critic class Model(object): def __init__(self, state_dim, action_dim, optimizer=None, actor_learning_rate=1e-4, critic_learning_rate=1e-3, tau = 0.001, sess=None): self.state_dim = state_dim self.action_dim = action_dim self.actor_learning_rate = actor_learning_rate self.critic_learning_rate = critic_learning_rate self.tau = tau #tf.reset_default_graph() self.sess = sess or tf.Session() self.global_step = tf.Variable(0, name="global_step", trainable=False) global_step_vars = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope="global_step") self.sess.run(tf.variables_initializer(global_step_vars)) self.actor_scope = "actor_net" with tf.name_scope(self.actor_scope): self.actor = DDPG_Actor(self.state_dim, self.action_dim, learning_rate=self.actor_learning_rate, tau=self.tau, scope=self.actor_scope, sess=self.sess) self.critic_scope = "critic_net" with tf.name_scope(self.critic_scope): self.critic = DDPG_Critic(self.state_dim, self.action_dim, learning_rate=self.critic_learning_rate, tau=self.tau, scope=self.critic_scope, sess=self.sess) def update(self, state_batch, action_batch, y_batch, sess=None): sess = sess or self.sess self.critic.update_source_critic_net(state_batch, action_batch, y_batch, sess) action_batch_for_grad = self.actor.predict_action_source_net(state_batch, sess) action_grad_batch = self.critic.get_action_grads(state_batch, action_batch_for_grad, sess) self.actor.update_source_actor_net(state_batch, action_grad_batch, sess) self.critic.update_target_critic_net(sess) self.actor.update_target_actor_net(sess) def predict_action(self, observation, sess=None): sess = sess or self.sess return self.actor.predict_action_source_net(observation, sess) if __name__ == '__main__': import numpy as np state_dim = 40 action_dim = 3 actor_learning_rate = np.random.rand(1) print("actor_learning_rate: ", actor_learning_rate) critic_learning_rate = np.random.rand(1) print("critic_learning_rate: ", critic_learning_rate) tau = np.random.rand(1) print("tau: ", tau) sess = tf.Session() model = Model(state_dim, action_dim, tau=tau, actor_learning_rate=actor_learning_rate[0], critic_learning_rate=critic_learning_rate[0], sess=sess) random_state = np.random.normal(size=state_dim) print("random_state", random_state) random_action = np.random.random(size=action_dim) print("random_action", random_action) # check prediction pred_action = model.predict_action(random_state) print("predict_action", pred_action) # check forward target_q = model.critic.predict_q_target_net([random_state], [random_action], sess) print("predict target q", target_q) y = target_q[0] + 1 model.update([random_state], [random_action], [y])
{ "alphanum_fraction": 0.6378258458, "author": null, "avg_line_length": 37.175257732, "converted": null, "ext": "py", "file": null, "hexsha": "d46e6e9beb440dccb73d0170bdc39dee3e2403fc", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 30, "max_forks_repo_forks_event_max_datetime": "2021-09-15T06:33:55.000Z", "max_forks_repo_forks_event_min_datetime": "2017-09-21T00:49:02.000Z", "max_forks_repo_head_hexsha": "7176d7fa5cbc20d3d31e9c01f6c1424bd3501ecc", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "code-cultivater/TensorAgent", "max_forks_repo_path": "model/ddpg_model.py", "max_issues_count": 2, "max_issues_repo_head_hexsha": "7176d7fa5cbc20d3d31e9c01f6c1424bd3501ecc", "max_issues_repo_issues_event_max_datetime": "2020-03-12T16:23:45.000Z", "max_issues_repo_issues_event_min_datetime": "2019-10-14T03:22:55.000Z", "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "code-cultivater/TensorAgent", "max_issues_repo_path": "model/ddpg_model.py", "max_line_length": 98, "max_stars_count": 55, "max_stars_repo_head_hexsha": "7176d7fa5cbc20d3d31e9c01f6c1424bd3501ecc", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "lightaime/TensorAgent", "max_stars_repo_path": "model/ddpg_model.py", "max_stars_repo_stars_event_max_datetime": "2021-12-06T14:38:28.000Z", "max_stars_repo_stars_event_min_datetime": "2017-08-03T08:04:54.000Z", "num_tokens": 765, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 3606 }
[STATEMENT] lemma f_Exec_Stream_Acc_Output_drop: " f_Exec_Comp_Stream_Acc_Output k output_fun trans_fun xs c \<up> n = f_Exec_Comp_Stream_Acc_Output k output_fun trans_fun (xs \<up> n) ( f_Exec_Comp trans_fun (xs \<down> n \<odot>\<^sub>f k) c)" [PROOF STATE] proof (prove) goal (1 subgoal): 1. f_Exec_Comp_Stream_Acc_Output k output_fun trans_fun xs c \<up> n = f_Exec_Comp_Stream_Acc_Output k output_fun trans_fun (xs \<up> n) (f_Exec_Comp trans_fun (xs \<down> n \<odot> k) c) [PROOF STEP] by (simp add: f_Exec_Comp_Stream_Acc_Output_def f_shrink_def f_Exec_Stream_expand_aggregate_map_drop)
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": "AutoFocus-Stream_AF_Stream_Exec", "hexsha": null, "include": null, "lang": null, "length": 1, "llama_tokens": 245, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
import sys import networkx as nx import matplotlib import matplotlib.pyplot as plt import pygraphviz from networkx.drawing.nx_agraph import graphviz_layout graph = nx.Graph() def dtoCOM(a): global graph return nx.dijkstra_path_length(graph, a, "COM") for orbit in sys.stdin: A, B = [i.strip() for i in orbit.split(")")] graph.add_edge(A,B) redges = zip(nx.dijkstra_path(graph, "YOU", "SAN")+[0], [0]+nx.dijkstra_path(graph, "YOU", "SAN")) redges = list(redges)[1:-1] labls = { "YOU": "YO", "SAN": "Santa", "COM": "COM" } pos = nx.spectral_layout(graph) plt.subplot(1, 1, 1) nx.draw_networkx_nodes(graph, pos, alpha=0.5, with_labels = True, node_size=10) nx.draw_networkx_nodes(graph,pos, nodelist=nx.dijkstra_path(graph, "YOU", "SAN") , node_size=10, node_color='r' , alpha=0.8, with_labels = True) nx.draw_networkx_labels(graph, pos, labels=labls) nx.draw_networkx_edges(graph, pos, edge_color='b') nx.draw_networkx_edges(graph, pos, edgelist=redges, edge_color='r') plt.show()
{ "alphanum_fraction": 0.7054187192, "author": null, "avg_line_length": 29, "converted": null, "ext": "py", "file": null, "hexsha": "167ef5d44bcb082d0a7c1ca96f8beb1326ad4fc2", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "8c1e1a0766b067fbe282dd482bc258275c5a3364", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "DVRodri8/advent-of-code-2019", "max_forks_repo_path": "6/draw.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "8c1e1a0766b067fbe282dd482bc258275c5a3364", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "DVRodri8/advent-of-code-2019", "max_issues_repo_path": "6/draw.py", "max_line_length": 144, "max_stars_count": null, "max_stars_repo_head_hexsha": "8c1e1a0766b067fbe282dd482bc258275c5a3364", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "DVRodri8/advent-of-code-2019", "max_stars_repo_path": "6/draw.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 309, "path": null, "reason": "import networkx,from networkx", "repo": null, "save_path": null, "sha": null, "size": 1015 }
import math from typing import Iterable, Union import numpy as np from shor.errors import CircuitError from shor.layers import _Layer from shor.utils.collections import flatten QbitOrIterable = Union[int, Iterable] class _Gate(_Layer): """Abstract base quantum gate class # Properties input_length = valid length of input qubits qubits = indices of qubits, to be used as input to gate. """ @property def symbol(self): return self.__class__.__name__.lower() def __init__(self, *qbits: QbitOrIterable, **kwargs): super().__init__(**kwargs) self.qbits = flatten(qbits) if qbits else [0] self.dimension = kwargs.get("dimension", 1) assert all(map(lambda q: type(q) == int, self.qbits)), str(self.qbits) try: assert len(self.qbits) % self.dimension == 0 except AssertionError: raise CircuitError( f"The input qbits length {len(self.qbits)} is not divisible by the '{self.symbol}' " f"gate's dimension {self.dimension}" ) @property def qubits(self): return self.qbits def to_gates(self): if len(self.qbits) > self.dimension: return [ self.__class__(self.qbits[i : i + self.dimension]) for i in range(0, len(self.qbits), self.dimension) ] return [self] @property def num_states(self): return np.power(2, self.dimension) def to_matrix(self) -> np.ndarray: return np.eye(self.num_states()) @property def matrix(self): return self.to_matrix() def invert(self): return self @property def I(self): return self.invert() def __invert__(self): return self.invert() class CNOT(_Gate): symbol = "CX" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 2 if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 0, 1], [0, 0, 1, 0]]) class CY(_Gate): symbol = "CY" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 2 if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 0, -1 * 1j], [0, 0, 1j, 0]]) class CSWAP(_Gate): symbol = "CSWAP" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 3 if not qubits: qubits = [0, 1, 2] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: cswap_matrix = np.eye(8) cswap_matrix[:, [5, 6]] = cswap_matrix[:, [6, 5]] return cswap_matrix class Hadamard(_Gate): symbol = "H" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.multiply(np.divide(1, np.sqrt(self.num_states)), np.array([[1, 1], [1, -1]])) class PauliX(_Gate): symbol = "X" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[0, 1], [1, 0]]) class PauliY(_Gate): symbol = "Y" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[0, -1j], [1j, 0]]) class PauliZ(_Gate): symbol = "Z" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0], [0, -1]]) class QFT(_Gate): def __init__(self, *qubits, **kwargs): if not qubits: qubits = [0, 1] super().__init__(*qubits, dimension=len(qubits), **kwargs) # def to_gates(self): # # TODO: translate this gate to base gates / CNOTs # pass def get_nth_unity_root(self, k): return np.exp((2j * np.pi * k) / self.num_states) def to_matrix(self) -> np.ndarray: m = np.array(np.ones((self.num_states, self.num_states)), dtype="complex") for i in range(1, self.num_states): for j in range(i, self.num_states): w = self.get_nth_unity_root(i * j) m[i, j] = w m[j, i] = w return np.around(np.multiply(1 / np.sqrt(self.num_states), m), decimals=15) class SWAP(_Gate): symbol = "SWAP" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 2 if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0, 0, 0], [0, 0, 1, 0], [0, 1, 0, 0], [0, 0, 0, 1]]) class Cx(_Gate): symbol = "CX" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 2 if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0, 0, 0], [0, 0, 0, 1], [0, 0, 1, 0], [0, 1, 0, 0]]) class CCNOT(_Gate): symbol = "CCX" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 3 if not qubits: qubits = [0, 1, 2] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array( [ [1, 0, 0, 0, 0, 0, 0, 0], [0, 1, 0, 0, 0, 0, 0, 0], [0, 0, 1, 0, 0, 0, 0, 0], [0, 0, 0, 1, 0, 0, 0, 0], [0, 0, 0, 0, 1, 0, 0, 0], [0, 0, 0, 0, 0, 1, 0, 0], [0, 0, 0, 0, 0, 0, 0, 1], [0, 0, 0, 0, 0, 0, 1, 0], ] ) class CRZ(_Gate): symbol = "CRZ" def __init__(self, *qubits, angle=0, **kwargs): kwargs["dimension"] = 2 self.angle = angle if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array( [ [1, 0, 0, 0], [0, 1, 0, 0], [0, 0, np.exp(-1j * self.angle / 2), 0], [0, 0, 0, np.exp(1j * self.angle / 2)], ] ) class CH(_Gate): symbol = "CH" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 2 if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array( [ [1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1 / np.sqrt(2), 1 / np.sqrt(2)], [0, 0, 1 / np.sqrt(2), -1 / np.sqrt(2)], ] ) class S(_Gate): symbol = "S" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0], [0, 1j]]) class Sdg(_Gate): symbol = "Sdg" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0], [0, -1j]]) class T(_Gate): symbol = "T" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0], [0, np.exp(1j * np.pi / 4)]]) class Tdg(_Gate): symbol = "Tdg" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0], [0, np.exp(-1j * np.pi / 4)]]) class ID(_Gate): symbol = "I" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0], [0, 1]]) class U1(_Gate): symbol = "U1" def __init__(self, *qubits, angle=0, **kwargs): kwargs["dimension"] = 1 self.angle = angle if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array([[1, 0], [0, np.exp(1j * self.angle)]]) class Cz(_Gate): symbol = "CZ" def __init__(self, *qubits, **kwargs): kwargs["dimension"] = 2 if not qubits: qubits = [0, 1] super().__init__(*qubits, **kwargs) @staticmethod def to_matrix() -> np.ndarray: return np.array([[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, -1]]) class Rx(_Gate): symbol = "RX" def __init__(self, *qubits, angle=math.pi / 2, **kwargs): kwargs["dimension"] = 1 self.angle = angle if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array( [ [math.cos(self.angle / 2), -math.sin(self.angle / 2) * 1j], [-math.sin(self.angle / 2) * 1j, math.cos(self.angle / 2)], ] ) class Ry(_Gate): symbol = "RY" def __init__(self, *qubits, angle=math.pi / 2, **kwargs): kwargs["dimension"] = 1 self.angle = angle if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array( [ [math.cos(self.angle / 2), -math.sin(self.angle / 2)], [math.sin(self.angle / 2), math.cos(self.angle / 2)], ] ) class Rz(_Gate): symbol = "RZ" def __init__(self, *qubits, angle, **kwargs): self.angle = angle kwargs["dimension"] = 1 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array( [[np.exp(-(1 / 2) * 1j * self.angle), 0], [0, np.exp((1 / 2) * 1j * self.angle)]], dtype="complex" ) class U3(_Gate): symbol = "U3" def __init__(self, *qubits, theta=0, phi=0, lam=0, **kwargs): kwargs["dimension"] = 1 self.theta = theta self.phi = phi self.lam = lam if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array( [ [math.cos(self.theta / 2), -np.exp(1j * self.lam) * math.sin(self.theta / 2)], [ np.exp(1j * self.phi) * math.sin(self.theta / 2), np.exp(1j * (self.phi + self.lam)) * math.cos(self.theta / 2), ], ] ) class U2(U3): symbol = "U2" def __init__(self, *qubits, phi=0, lam=0, **kwargs): super().__init__(*qubits, theta=np.pi / 2, phi=phi, lam=lam, **kwargs) self.symbol = "u2" class Init_x(_Gate): def __init__(self, *qubits, **kwargs): self.H = Hadamard(0) kwargs["dimension"] = 1 super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return self.H.to_matrix() class Init_y(_Gate): def __init__(self, *qubits, **kwargs): self.H = Hadamard(0) self.S = S() kwargs["dimension"] = 1 super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return self.S.to_matrix().dot(self.H.to_matrix()) class Cr(_Gate): symbol = "CU1" def __init__(self, *qubits, angle, **kwargs): self.angle = angle kwargs["dimension"] = 2 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array([[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, np.exp(1j * self.angle)]], dtype="complex") class CRk(_Gate): def __init__(self, *qubits, k, **kwargs): self.k = k kwargs["dimension"] = 2 if not qubits: qubits = [0] super().__init__(*qubits, **kwargs) def to_matrix(self) -> np.ndarray: return np.array( [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, np.exp(2 * 1j * np.pi / 2 ** self.k)]], dtype="complex" ) # Aliases H = h = Hadamard X = x = PauliX Y = y = PauliY Z = z = PauliZ swap = SWAP Fredkin = cswap = CSWAP CX = cx = CNOT
{ "alphanum_fraction": 0.5080947384, "author": null, "avg_line_length": 23.9964028777, "converted": null, "ext": "py", "file": null, "hexsha": "696bfe59953f2dc657d64439260348f8fed03c23", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "6dda779779bc425ef7a726fc427620924255050e", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "vanmagnan/shor", "max_forks_repo_path": "shor/gates.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "6dda779779bc425ef7a726fc427620924255050e", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "vanmagnan/shor", "max_issues_repo_path": "shor/gates.py", "max_line_length": 120, "max_stars_count": null, "max_stars_repo_head_hexsha": "6dda779779bc425ef7a726fc427620924255050e", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "vanmagnan/shor", "max_stars_repo_path": "shor/gates.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 4102, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 13342 }
import argparse import logging import json from pprint import pprint import numpy as np import os from astropy.coordinates import EarthLocation, SkyCoord from astropy.time import Time from astropy import units as u from astropy.coordinates import AltAz from datetime import datetime import ctbend.ctbendbase.CTBend as CTBend from ctbend.ctbendtrainer.CTBendGeometry import XYZVector, e_phi, e_theta from ctbend.ctbendbase.PointingData import CCDCoordinate, DriveCoordinate,\ PointingData, PointingDataset logger = logging.getLogger() logging.basicConfig(level=logging.INFO) class MCSetup(object): def __init__(self, config_dict): self.config_dict = config_dict def random_star(self): """Return a random star coordinate. """ ra_h = np.random.uniform(0, 24) ra_m = np.random.uniform(0, 60) ra_s = np.random.uniform(0, 60) star_ra = str(int(ra_h)) + "h" star_ra += str(int(ra_m)) + "m" star_ra += str(int(ra_s)) + "s" dec_d = np.random.uniform(0, 90) dec_m = np.random.uniform(0, 60) dec_s = np.random.uniform(0, 60) star_dec = str(int(dec_d)) + "d" star_dec += str(int(dec_m)) + "m" star_dec += str(int(dec_s)) + "s" return SkyCoord(ra=star_ra, dec=star_dec) @property def location(self): """Observation location. """ lat = u.Quantity(self.config_dict["location_lat"]) lon = u.Quantity(self.config_dict["location_lon"]) height = u.Quantity(self.config_dict["location_height"]) return EarthLocation(lat=lat, lon=lon, height=height) @property def telescope_focal_length(self): """Focal length of the telescope as quantity with unit. """ return u.Quantity(self.config_dict["telescope_focal_length"]) @property def ccd_focal_length(self): """CCD focal length as quantity with unit. """ return u.Quantity(self.config_dict["ccd_focal_length"]) @property def ccd_pixel_size(self): """Size of a CCD pixel (assumed to be a square) as quantity with unit. """ return u.Quantity(self.config_dict["ccd_pixel_size"]) @property def delta_t(self): """Time between two CCD images as quantity with unit. """ return u.Quantity(self.config_dict["delta_t"]) @property def n_tracking(self): """Number of tracking points. """ return int(self.config_dict["n_tracking"]) @property def start_timestamp(self): """Timestamp at start of tracking. """ return self.config_dict["start_timestamp"] def tracking_timestamps(self): """Yields a tracking timestamp. """ delta_t = u.Quantity(self.config_dict["delta_t"]) start_timestamp = self.config_dict["start_timestamp"] for i in range(int(self.config_dict["n_tracking"])): timestamp = start_timestamp + i * delta_t.to(u.s).value yield timestamp @property def pixel_scale(self): """CCD pixel scale in radians. """ scale = self.ccd_pixel_size.to(u.m).value scale /= self.ccd_focal_length.to(u.m).value scale *= u.rad return scale @property def bending(self): """Returns the bending model applied during data-taking. """ parameters = self.config_dict["bending"]["parameters"] parameters0 = parameters return getattr(CTBend, self.config_dict["bending"]["model"])(parameters0) def measured_x1x2(self, true_x1, true_x2): """Position of star in CCD coordinates. """ sigma = self.config_dict["sigma"]["size"] sigma2 = np.power(sigma, 2) (dx1, dx2) = np.random.multivariate_normal([0, 0], cov=[[sigma2, 0], [0, sigma2]]) offset_x1 = float(self.config_dict["sigma"]["offset_x1"]) offset_x2 = float(self.config_dict["sigma"]["offset_x2"]) return (true_x1 + dx1 + offset_x1, true_x2 + dx2 + offset_x2) if __name__ == "__main__": parser = argparse.ArgumentParser() parser.add_argument("--config", type=str, help="MC config json file", dest="CONFIG", required=True) parser.add_argument("--outfile", type=str, help="MC tracking data outfile", dest="OUTFILE", required=True) options = parser.parse_args() if os.path.isfile(options.OUTFILE): raise RuntimeError(options.OUTFILE + " already exists") with open(options.CONFIG) as fin: config = json.load(fin) pprint(config) setup = MCSetup(config) info = "Pixelscale: " + str(setup.pixel_scale.to(u.arcsec).value) info += " arcsec" logger.info(info) pointing_model_for_data_taking = CTBend.ConstantOffsetModel( parameters={"azimuth_offset_deg": 0., "elevation_offset_deg": 0.}) pointing_dataset = PointingDataset( pixelscale=setup.pixel_scale.to(u.arcsec).value, pointing_model=pointing_model_for_data_taking.serialize()) for timestamp in setup.tracking_timestamps(): def random_star_altaz(minimal_altitude_deg: float = 5): """Returns random altaz coordinates for a test star. Args: minimal_altitude_deg: Minimal altitude of the star in degrees. """ def get_random_star(): time = datetime.fromtimestamp(timestamp) observing_time = Time(time) aa = AltAz(location=setup.location, obstime=observing_time) return setup.random_star().transform_to(aa) star_altaz = get_random_star() while star_altaz.alt.to(u.deg).value < minimal_altitude_deg: star_altaz = get_random_star() return star_altaz star_altaz = random_star_altaz() bending = setup.bending az_star_deg = star_altaz.az.to(u.deg).value el_star_deg = star_altaz.alt.to(u.deg).value delta_az_deg = bending.delta_azimuth(az=az_star_deg, el=el_star_deg) delta_el_deg = bending.delta_elevation(az=az_star_deg, el=el_star_deg) telescope_az = (az_star_deg - delta_az_deg) * u.deg telescope_el = (el_star_deg - delta_el_deg) * u.deg telescope = XYZVector(az=telescope_az.to(u.deg).value, el=telescope_el.to(u.deg).value) def get_e_phi(): delta_az_derivative_phi = bending.delta_azimuth_derivative_phi( az=telescope_az.to(u.deg).value, el=telescope_el.to(u.deg).value) delta_el_derivative_phi = bending.delta_elevation_derivative_phi( az=telescope_az.to(u.deg).value, el=telescope_el.to(u.deg).value) _e_phi = e_phi(az_tel=telescope.az, el_tel=telescope.alt, delta_az_derivative_phi=delta_az_derivative_phi, delta_el_derivative_phi=delta_el_derivative_phi) return _e_phi def get_e_theta(): delta_az_derivative_theta = bending.delta_azimuth_derivative_theta( az=telescope_az.to(u.deg).value, el=telescope_el.to(u.deg).value) delta_el_derivative_theta = \ bending.delta_elevation_derivative_theta( az=telescope_az.to(u.deg).value, el=telescope_el.to(u.deg).value) _e_theta = e_theta( az_tel=telescope.az, el_tel=telescope.alt, delta_az_derivative_theta=delta_az_derivative_theta, delta_el_derivative_theta=delta_el_derivative_theta) return _e_theta def telescope_ccd_position(): """The pointing direction of the telescope in CCD coordinates is the center of the LED pattern. By definition, this is the zero-vector in CCD coordinates in this MC. """ x1_tel = 0 x2_tel = 0 return CCDCoordinate(x1_tel, x2_tel) def star_ccd_position(): star_vector = XYZVector(star_altaz.az.to(u.deg).value, star_altaz.alt.to(u.deg).value) focal_length_mm = setup.ccd_focal_length.to(u.mm).value image_star_length = focal_length_mm / (telescope * star_vector) image = telescope * (star_vector * telescope) * 2. - star_vector image_star = image * image_star_length fp_u = image_star * get_e_phi() fp_v = image_star * get_e_theta() x1_star = fp_u / setup.ccd_pixel_size.to(u.mm).value x2_star = fp_v / setup.ccd_pixel_size.to(u.mm).value measured_x1_star, measured_x2_star = setup.measured_x1x2(x1_star, x2_star) return CCDCoordinate(measured_x1_star, measured_x2_star) drive_position = DriveCoordinate(telescope.az, telescope.alt) pointing_data = PointingData(telescope=telescope_ccd_position(), star=star_ccd_position(), drive_position=drive_position) pointing_dataset.append(pointing_data) logger.debug("Pointing data:") logger.debug(pointing_data) pointing_dataset.save(options.OUTFILE)
{ "alphanum_fraction": 0.5727593646, "author": null, "avg_line_length": 34.2214765101, "converted": null, "ext": "py", "file": null, "hexsha": "6dabd5f28654c086b3d4d43656607b582aa8a14a", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "826c147a9b1830e66ff63fa356e7fd48b7615c96", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "residualsilence/ctbend", "max_forks_repo_path": "ctbendtrainer/test/trackingmc.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "826c147a9b1830e66ff63fa356e7fd48b7615c96", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "residualsilence/ctbend", "max_issues_repo_path": "ctbendtrainer/test/trackingmc.py", "max_line_length": 79, "max_stars_count": null, "max_stars_repo_head_hexsha": "826c147a9b1830e66ff63fa356e7fd48b7615c96", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "residualsilence/ctbend", "max_stars_repo_path": "ctbendtrainer/test/trackingmc.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2135, "path": null, "reason": "import numpy,from astropy", "repo": null, "save_path": null, "sha": null, "size": 10198 }
[STATEMENT] lemma dj_cp: fixes pi1::"'x prm" and pi2::"'y prm" and x ::"'a" assumes cp: "cp TYPE ('a) TYPE('x) TYPE('y)" and dj: "disjoint TYPE('y) TYPE('x)" shows "pi1\<bullet>(pi2\<bullet>x) = (pi2)\<bullet>(pi1\<bullet>x)" [PROOF STATE] proof (prove) goal (1 subgoal): 1. pi1 \<bullet> pi2 \<bullet> x = pi2 \<bullet> pi1 \<bullet> x [PROOF STEP] by (simp add: cp1[OF cp] dj_perm_perm_forget[OF dj])
{ "alphanum_fraction": null, "author": null, "avg_line_length": null, "converted": null, "ext": null, "file": null, "hexsha": null, "include": null, "lang": null, "length": 1, "llama_tokens": 201, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": null, "max_forks_repo_licenses": null, "max_forks_repo_name": null, "max_forks_repo_path": null, "max_issues_count": null, "max_issues_repo_head_hexsha": null, "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": null, "max_issues_repo_name": null, "max_issues_repo_path": null, "max_line_length": null, "max_stars_count": null, "max_stars_repo_head_hexsha": null, "max_stars_repo_licenses": null, "max_stars_repo_name": null, "max_stars_repo_path": null, "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": null, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": null }
""" To run tests: pytest test_game.py """ import numpy as np import curling.constants import log_handler from curling import board as board_utils from curling import constants as c from curling import game from curling import simulation from curling import utils log_handler.flush_on_error() def test_simulation_setupBoard_0(): sim = simulation.Simulation() stones = list(sim.getStones()) assert len(stones) == 0 def test_simulation_setupBoard_1(): curl = game.CurlingGame() board = curl.getInitBoard() sim = curl.sim sim.setupBoard(board) sim.addStone(c.P1_COLOR, 0, utils.TEE_LINE) sim.addStone(c.P2_COLOR, 0, utils.TEE_LINE) stones = list(sim.getStones()) assert len(stones) == 2 def test_simulation_getNextStoneId(): curl = game.CurlingGame() r = utils.STONE_RADIUS i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 0 # for red curl.sim.addStone(c.P1_COLOR, 0, utils.HOG_LINE) i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 0 # for blue curl.sim.addStone(c.P2_COLOR, 2 * r, utils.HOG_LINE + 2 * r) i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 1 # for red curl.sim.addStone(c.P1_COLOR, 4 * r, utils.HOG_LINE + 4 * r) i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 1 # for blue def test_simulation_getNextStoneId_with_removed(): curl = game.CurlingGame() curl.sim.addShooterAsInvalid() curl.sim.addShooterAsInvalid() curl.sim.addShooterAsInvalid() curl.sim.addShooterAsInvalid() r = utils.STONE_RADIUS i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 2 # for red curl.sim.addStone(c.P1_COLOR, 0, utils.HOG_LINE) i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 2 # for blue curl.sim.addStone(c.P2_COLOR, 2 * r, utils.HOG_LINE + 2 * r) i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 3 # for red curl.sim.addStone(c.P1_COLOR, 4 * r, utils.HOG_LINE + 4 * r) i = simulation.getNextStoneId(curl.sim.getBoard()) assert i == 3 # for blue def test_simulation_getBoard_is_symmetric(): """Create a board then convert it to simulation and back to board.""" curl = game.CurlingGame() expected = curl.sim.getBoard() curl.sim.addStone(c.P1_COLOR, 0, utils.TEE_LINE) curl.sim.setupBoard(expected) actual = curl.sim.getBoard() actual_position = list(board_utils.get_xy_team1(actual)) expected_position = list(board_utils.get_xy_team1(expected)) np.testing.assert_array_equal(actual_position, expected_position) def SKIP_test_simulation_getBoard_button(): sim = game.CurlingGame().sim board = sim.getBoard() board[-1][0] = c.EMPTY button_x, button_y = curling.constants.BUTTON_POSITION board_x, board_y = utils.realToBoard(button_x, button_y) board[board_x][board_y] = c.P1 sim.setupBoard(board) stones = sim.getStones() assert len(stones) == 1 stone_x, stone_y = stones[0].body.position assert (button_x, button_y) == (stone_x, stone_y) recalculated_board = sim.getBoard() expected = np.argwhere(board == c.P1) actual = np.argwhere(recalculated_board == c.P1) np.testing.assert_array_equal(expected, actual) def SKIP_test_simulation_getBoard_right_edge(): sim = game.CurlingGame().sim board = sim.getBoard() board[-1][0] = c.EMPTY button_x, button_y = curling.constants.BUTTON_POSITION button_x -= utils.dist(inches=10) # NOTE - shifting to the left board_x, board_y = utils.realToBoard(button_x, button_y) board[board_x][board_y] = c.P1 sim.setupBoard(board) stones = sim.getStones() assert len(stones) == 1 stone_x, stone_y = stones[0].body.position assert (button_x, button_y) == (stone_x, stone_y) recalculated_board = sim.getBoard() expected = np.argwhere(board == c.P1) actual = np.argwhere(recalculated_board == c.P1) np.testing.assert_array_equal(expected, actual) def test_invalid_shooter(): curl = game.CurlingGame() curl.sim.addShooterAsInvalid() def test_simulation_getBoard_has_extra_data(): curl = game.CurlingGame() expected = curl.sim.getBoard() board_utils.configure_hammer_2_scenario(expected) curl.sim.setupBoard(expected) actual = curl.sim.getBoard() # Not called during setup() because useless board_utils.update_distance_and_score(expected) np.testing.assert_array_equal(actual, expected)
{ "alphanum_fraction": 0.700264667, "author": null, "avg_line_length": 26.514619883, "converted": null, "ext": "py", "file": null, "hexsha": "06fcbe0272354b4f01046abddb88d8329b17ee08", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 1, "max_forks_repo_forks_event_max_datetime": "2021-08-23T02:17:06.000Z", "max_forks_repo_forks_event_min_datetime": "2021-08-23T02:17:06.000Z", "max_forks_repo_head_hexsha": "1b2641ef823a8dedd49159ecd5b69b77c7165731", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "mikhail/alpha-zero-general", "max_forks_repo_path": "curling/test_simulation.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "1b2641ef823a8dedd49159ecd5b69b77c7165731", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "mikhail/alpha-zero-general", "max_issues_repo_path": "curling/test_simulation.py", "max_line_length": 73, "max_stars_count": null, "max_stars_repo_head_hexsha": "1b2641ef823a8dedd49159ecd5b69b77c7165731", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "mikhail/alpha-zero-general", "max_stars_repo_path": "curling/test_simulation.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1200, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 4534 }
# Imports here import matplotlib.pyplot as plt import numpy as np import time import json import torch from torch import nn from torch import optim import torch.nn.functional as F from torchvision import datasets, transforms, models from collections import OrderedDict import PIL from PIL import Image import argparse def get_training_input_args(): model_list = ['vgg16','vgg19','vgg13','densenet','alexnet', 'densenet201', 'resnet18'] parser = argparse.ArgumentParser(description='Train and save a model checkpoint') parser.add_argument('data_dir', type = str, help = 'path to dataset folder') parser.add_argument('--save_dir', type = str, help = 'Directory where the checkpoint will be saved', default= 'checkpoint.pth') parser.add_argument('--arch', type = str, help = 'Model Architecture eg vgg,densenet, alexnet, resnet18', default= 'vgg13', choices = model_list) parser.add_argument('--learning_rate', type = float, help = 'Learning Rate', default= 0.01) parser.add_argument('--hidden_units', type = int, help = 'List of number of nodes in hidden layers', nargs='+', default= [1000]) parser.add_argument('--epochs', type = int, help = 'Number of epochs', default = 20) parser.add_argument('--gpu', help = 'Switch to gpu ', action= 'store_true') parser.add_argument('--dropout', type = float, help = 'Dropout for training', default = 0.5) in_arg = parser.parse_args() return in_arg def get_prediction_input_args(): parser = argparse.ArgumentParser(description='Predict the category of a flower') parser.add_argument('image_path', type = str, help = 'path to flower whose class is to be predicted') parser.add_argument('checkpoint', type = str, help = 'Directory where the checkpoint was saved' ) parser.add_argument('--top_k', type = int, help = 'number of most likely classes', default= 3) parser.add_argument('--category_names', help = 'Enter JSON file', default= 'cat_to_name.json') parser.add_argument('--gpu', help = 'Switch to gpu ', action= 'store_true') in_arg = parser.parse_args() return in_arg def check_command_line_args_prediction(in_arg): print("______________________________________") print("____ Arguments used for prediction ___") print("______________________________________") print("\n image_path =", in_arg.image_path, "\n checkpoint =", in_arg.checkpoint, "\n top_k =", in_arg.top_k, "\n category_names =", in_arg.category_names, "\n gpu =", in_arg.gpu) def check_command_line_args(in_arg): print("______________________________________") print("____ Arguments used for training ___") print("______________________________________") print("\n data_dir =", in_arg.data_dir, "\n arch =", in_arg.arch, "\n save_dir =", in_arg.save_dir, "\n dropout =", in_arg.dropout, "\n learning_rate =", in_arg.learning_rate, "\n hidden_units =", in_arg.hidden_units, "\n epochs =", in_arg.epochs, "\n gpu =", in_arg.gpu) # Loading the pretrained Network in_features = 0 def load_model(model_name): if 'vgg' in model_name: model = models.__dict__[model_name](pretrained=True) in_features = model.classifier[0].in_features elif model_name == 'alexnet': model = models.alexnet(pretrained = True) in_features = 9216 elif 'densenet' in model_name: model = models.__dict__[model_name](pretrained=True) in_features = model.classifier.in_features return model, in_features # Building the Network class FlowerNetwork(nn.Module): def __init__(self, input_size, output_size, hidden_layers, drop_p=0.5): super().__init__() self.hidden_layers=nn.ModuleList([nn.Linear(input_size, hidden_layers[0])]) layer_sizes = zip(hidden_layers[:-1], hidden_layers[1:]) self.hidden_layers.extend([nn.Linear(h1, h2) for h1, h2 in layer_sizes]) self.output = nn.Linear(hidden_layers[-1], output_size) self.dropout = nn.Dropout(p=drop_p) def forward(self, x): for linear in self.hidden_layers: x = F.relu(linear(x)) x = self.dropout(x) x = self.output(x) return F.log_softmax(x, dim=1) def validation(model, device, dataloaders_valid, criterion): test_loss = 0 accuracy = 0 model.to(device) for images, labels in dataloaders_valid: images = images.to(device) labels = labels.to(device) output = model.forward(images) test_loss += criterion(output, labels).item() # Calculating the accuracy # take exponents to get the probabilities ps = torch.exp(output) equality = (labels.data == output.max(dim=1)[1]) accuracy += equality.type(torch.FloatTensor).mean() return test_loss, accuracy def trainer(device, model, dataloaders, print_every,criterion,optimizer,epochs, dataloaders_valid): steps = 0 running_loss = 0 model.to(device) for e in range(epochs): model.train() for images, labels in dataloaders: steps += 1 images = images.to(device) labels = labels.to(device) optimizer.zero_grad() # forward step happens here output = model.forward(images) loss = criterion(output, labels) #backpropagation step loss.backward() optimizer.step() running_loss += loss.item() if steps % print_every == 0: model.eval() with torch.no_grad(): test_loss, accuracy = validation(model, device, dataloaders_valid, criterion) print("Epoch: {}/{}.. ".format(e+1, epochs), "Training Loss: {:.3f}.. ".format(running_loss/print_every), "Validation Loss: {:.3f}.. ".format(test_loss/len(dataloaders_valid)), " Accuracy: {:.3f}".format(accuracy/len(dataloaders_valid))) running_loss = 0 model.train() print("__Training successfully completed__") # TODO: Do validation on the test set def test_network(model, device,dataloaders_test): model.to(device) model.eval() accurately_classified_count = 0 total = 0 for images, labels in dataloaders_test: images = images.to(device) labels = labels.to(device) output = model(images) _, prediction = torch.max(output.data,1) total += labels.size(0) accurately_classified_count += torch.sum(prediction == labels.data).item() testing_accuracy = 100 * (accurately_classified_count/total) return testing_accuracy # TODO: Save the checkpoint def save_checkpoint(model,model_name, filename, image_datasets, optimizer): model.class_to_idx = image_datasets.class_to_idx checkpoint = {'model_name': model_name, 'classifier': model.classifier, 'state_dict': model.state_dict(), 'optimizer' : optimizer.state_dict(), 'class_to_idx': model.class_to_idx } torch.save(checkpoint, filename ) print("Checkpoint successfully saved to: ", filename) def load_checkpoint(filename,device): checkpoint = torch.load(filename) model_name = checkpoint['model_name'] model = models.__getattribute__(model_name)(pretrained = True) model.to(device) for param in model.parameters(): param.requires_grad = False model.classifier = checkpoint['classifier'] model.class_to_idx = checkpoint['class_to_idx'] model.load_state_dict(checkpoint['state_dict']) optimizer = checkpoint['optimizer'] model.eval() return (model) def process_image(image): ''' Scales, crops, and normalizes a PIL image for a PyTorch model, returns an Numpy array ''' # TODO: Process a PIL image for use in a PyTorch model img = Image.open(image) if img.width > img.height: ratio = img.width/img.height img = img.resize(size=( int(round(ratio*256,0)),256)) else: ratio = img.height/img.width img = img.resize(size=(256, int(round(ratio*256,0)))) bottom = (img.height + 224)/2 right = (img.width + 224)/2 top = (img.height - 224)/2 left = (img.width - 224)/2 img = img.crop(box=(left,top,right,bottom)) np_image = np.array(img)/255 mean = np.array([0.485, 0.456, 0.406]) std = np.array([0.229, 0.224, 0.225]) img = (np_image - mean)/std img = img.transpose((2, 0, 1)) return img def imshow(image, ax=None, title=None): """Imshow for Tensor.""" if ax is None: fig, ax = plt.subplots() # PyTorch tensors assume the color channel is the first dimension # but matplotlib assumes is the third dimension image = image.transpose((1, 2, 0)) # Undo preprocessing mean = np.array([0.485, 0.456, 0.406]) std = np.array([0.229, 0.224, 0.225]) image = std * image + mean # Image needs to be clipped between 0 and 1 or it looks like noise when displayed image = np.clip(image, 0, 1) ax.imshow(image) return ax def predict(image_path, model, device, category_names, topk=5): ''' Predict the class (or classes) of an image using a trained deep learning model. ''' # TODO: Implement the code to predict the class from an image file with open(category_names, 'r') as f: cat_to_name = json.load(f) image = process_image(image_path) image = torch.from_numpy(image).type(torch.FloatTensor) image = image.unsqueeze(0).float() image = image.to(device) model.eval() with torch.no_grad(): output = model.forward(image) ps = torch.exp(output) probs, indices = torch.topk(ps, topk) Probs = np.array(probs.data[0]) Indices = np.array(indices.data[0]) # invert class_to_idx idx_to_class = {idx:Class for Class,idx in model.class_to_idx.items()} classes = [idx_to_class[i] for i in Indices] labels = [cat_to_name[Class] for Class in classes] print("\n ___Most likely flower class with associated Probability___ ") print(labels[0], " : ", Probs[0]) print(":\n ___Top K classes along with associated probabilities ___: \n") for i, val in enumerate(Probs): print( labels[i], " : ", Probs[i]) return Probs,labels def duration(start_time, end_time): tot_time = end_time - start_time print("\n** Total Elapsed Runtime:", str(int((tot_time/3600))) + ":" + str(int((tot_time%3600)/60)) + ":" + str(int((tot_time%3600)%60)))
{ "alphanum_fraction": 0.6274171584, "author": null, "avg_line_length": 33.7883435583, "converted": null, "ext": "py", "file": null, "hexsha": "7c912d11c4873d7dc4f4c126e7652382860f22d6", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "7289adc5f1e811bf09144bd31dbb7c010a27dbc5", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "AineKiraboMbabazi/AIPND_Image_Classifier", "max_forks_repo_path": "helper_functions.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "7289adc5f1e811bf09144bd31dbb7c010a27dbc5", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "AineKiraboMbabazi/AIPND_Image_Classifier", "max_issues_repo_path": "helper_functions.py", "max_line_length": 149, "max_stars_count": null, "max_stars_repo_head_hexsha": "7289adc5f1e811bf09144bd31dbb7c010a27dbc5", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "AineKiraboMbabazi/AIPND_Image_Classifier", "max_stars_repo_path": "helper_functions.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 2591, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 11015 }
import os from os.path import basename os.environ['CUDA_VISIBLE_DEVICES'] = '' from keras.models import load_model import matplotlib.pyplot as plt from skimage import io import numpy as np from skimage.color import gray2rgb, label2rgb from skimage.io import imsave from datetime import datetime from functions import your_loss import glob from scipy.misc import imread #from skimage.io import imread import glob import re import time import random from datetime import datetime labels = 4 channels = 1 size = 56 inner_size = 36 ysize = 36 batch_size = 254 def output_to_colors(result, x): zeros = np.zeros((rows,cols,4)) #zeros[:,:,:-1]=gray2rgb(x.copy()) #zeros = gray2rgb(x.copy()) #output = result.argmax(axis=-1) zeros[output==2]=[0,0,1] return zeros def chunks(l, n): """Yield successive n-sized chunks from l.""" for i in range(0, len(l), n): yield l[i:i + n] def get_tiles(img, inner_size, overlap): img_padded = np.pad(img, ((overlap,overlap), (overlap,overlap)), mode='reflect') xs = [] for i in xrange(0, img.shape[0], inner_size): for j in xrange(0, img.shape[1], inner_size): #print(i-overlap+overlap,i+inner_size+overlap+overlap,j-overlap+overlap, j+inner_size+overlap+overlap) img_overlapped = img_padded[i:i+inner_size+overlap+overlap,j:j+inner_size+overlap+overlap] xs.append(img_overlapped) return xs def natural_sort(l): convert = lambda text: int(text) if text.isdigit() else text.lower() alphanum_key = lambda key: [ convert(c) for c in re.split('([0-9]+)', key) ] return sorted(l, key = alphanum_key) models = sorted(natural_sort(glob.glob('models/*')), key=lambda name: int(re.search(r'\d+', name).group()), reverse=True)[0:1] print(models) tick = datetime.now() for model_n in models: model = load_model(model_n, custom_objects={'your_loss': your_loss}) print("Loaded :%s", model_n) files_all = glob.glob('temp_cropped/*') #files_all = glob.glob('cleaned/raw/*.png') #files_all = glob.glob('images/single_frames/*.png') #files_all = glob.glob('images/all_grey/15_jpgs/*.jpg') #files_all = random.sample(files_all) #files_all = glob.glob('cleaned/penta/*') #files = files+files+files# + files[-4:-1] #print(files) file_chunks = chunks(files_all, batch_size) for idx, files in enumerate(file_chunks): file_names = [basename(path) for path in files] #print(file_names) imgs = np.array([np.pad(imread(fl, mode='L'), (8,8), mode='reflect').astype(float)/255 for fl in files]) #import pdb; pdb.set_trace() tiles = np.array([get_tiles(img, 36, 10) for img in imgs]) #print(file_chunks) #print("Processing: %s"%(fl)) #print("Imgs shape: %s", tiles.shape) #Create input tensor xs = tiles.reshape(imgs.shape[0]*len(tiles[0]),size,size,channels) #print(np.unique(xs[0])) start_time = time.time() # Predict output ys = model.predict(xs) print("---- %s seconds for size: %d ----"%(time.time()-start_time, xs.shape[0])) ys = ys.reshape(imgs.shape[0],len(tiles[0]), ysize, ysize, labels) # Stitch it together for ix,y in enumerate(ys): #imgcount = 0 count= 0 img = imgs[ix] tile_output = np.zeros((img.shape[0],img.shape[1],4)) zeros = np.zeros((img.shape[0],img.shape[1],4)) for i in xrange(0, img.shape[0], inner_size): for j in xrange(0, img.shape[1], inner_size): zeros[i:i+inner_size,j:j+inner_size] = y[count] count += 1 for i in range(img.shape[0]): for j in range(img.shape[1]): output = tile_output[i,j] zeros[i,j,np.argmax(output)] = 1 #count += 1 #import pdb; pdb.set_trace() zeros[:,:,3]=1 #color = output_to_colors(zeros, img) #colors = [output_to_colors(y, imgs[i]) for i,y in enumerate(ys)] #colors = [label2rgb(y.argmax(axis=-1), image=imgs[i], colors=[(1,0,0), (0,1,0), (0,0,1), (0,0,0)], alpha=0.9, bg_label=3) for i,y in enumerate(ys)] #[plt.imsave('plots/%s_%s'%(model_n, file_names[i]), zeros) for i,zeros in enumerate(colors)] #print(file_names) plt.imsave('plots/results/%s.png'%(file_names[ix]), zeros) print "total processing done in: "+str((datetime.now()-tick).total_seconds())
{ "alphanum_fraction": 0.597693786, "author": null, "avg_line_length": 34.947761194, "converted": null, "ext": "py", "file": null, "hexsha": "26c732c18116a026621aceeb43de31b4fecddbe6", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "44e7318ce6a974bd9fe975219dd3f45727bc9522", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "crackmech/fly-walk", "max_forks_repo_path": "output_multiple_crop.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "44e7318ce6a974bd9fe975219dd3f45727bc9522", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "crackmech/fly-walk", "max_issues_repo_path": "output_multiple_crop.py", "max_line_length": 164, "max_stars_count": null, "max_stars_repo_head_hexsha": "44e7318ce6a974bd9fe975219dd3f45727bc9522", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "crackmech/fly-walk", "max_stars_repo_path": "output_multiple_crop.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 1239, "path": null, "reason": "import numpy,from scipy", "repo": null, "save_path": null, "sha": null, "size": 4683 }
C C $Id: pakrsp.f,v 1.5 2008-07-27 00:17:10 haley Exp $ C C Copyright (C) 2000 C University Corporation for Atmospheric Research C All Rights Reserved C C The use of this Software is governed by a License Agreement. C REAL FUNCTION PAKRSP (ANG) C C Function to convert DMS packed angle into radians. C IMPLICIT REAL (A-Z) DATA SECRAD /0.4848136811095359E-5/ C C Convert angle to seconds of arc. C SEC = PAKSSP (ANG) C C Convert angle to radians. C PAKRSP = SEC * SECRAD C RETURN END
{ "alphanum_fraction": 0.648312611, "author": null, "avg_line_length": 20.8518518519, "converted": null, "ext": "f", "file": null, "hexsha": "a9ec52742119d6515683a05c9da27279a7c9935e", "include": null, "lang": "FORTRAN", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": 58, "max_forks_repo_forks_event_max_datetime": "2022-03-15T09:13:00.000Z", "max_forks_repo_forks_event_min_datetime": "2016-12-14T00:15:22.000Z", "max_forks_repo_head_hexsha": "a87114a689a1566e9aa03d85bcf6dc7325b47633", "max_forks_repo_licenses": [ "Apache-2.0" ], "max_forks_repo_name": "tenomoto/ncl", "max_forks_repo_path": "ncarg2d/src/libncarg/ezmapc/pakrsp.f", "max_issues_count": 156, "max_issues_repo_head_hexsha": "a87114a689a1566e9aa03d85bcf6dc7325b47633", "max_issues_repo_issues_event_max_datetime": "2022-03-30T07:02:21.000Z", "max_issues_repo_issues_event_min_datetime": "2017-09-22T09:56:48.000Z", "max_issues_repo_licenses": [ "Apache-2.0" ], "max_issues_repo_name": "tenomoto/ncl", "max_issues_repo_path": "ncarg2d/src/libncarg/ezmapc/pakrsp.f", "max_line_length": 62, "max_stars_count": 210, "max_stars_repo_head_hexsha": "a87114a689a1566e9aa03d85bcf6dc7325b47633", "max_stars_repo_licenses": [ "Apache-2.0" ], "max_stars_repo_name": "tenomoto/ncl", "max_stars_repo_path": "ncarg2d/src/libncarg/ezmapc/pakrsp.f", "max_stars_repo_stars_event_max_datetime": "2022-03-24T19:15:32.000Z", "max_stars_repo_stars_event_min_datetime": "2016-11-24T09:05:08.000Z", "num_tokens": 178, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 563 }
rng = StableRNG(614) # convert a Binary vector into vector of +1 or -1 values # (for testing only): pm1(y) = Int8(2) .* (Int8.(MLJBase.int(y))) .- Int8(3) const MARGIN_LOSSES = MLJBase.MARGIN_LOSSES const DISTANCE_LOSSES = MLJBase.DISTANCE_LOSSES # using `WeightedSum` instead of `WeightedMean`; see # https://github.com/JuliaML/LossFunctions.jl/issues/149 WeightedSum(w) = LossFunctions.AggMode.WeightedMean(w, normalize=false) @testset "naked" begin @test MLJBase.naked(MLJBase.LossFunctions.PeriodicLoss{Float64}) == :PeriodicLoss @test MLJBase.naked(MLJBase.LossFunctions.PeriodicLoss) == :PeriodicLoss end @testset "LossFunctions.jl - binary" begin y = categorical(["yes", "yes", "no", "yes"]) yes, no = y[1], y[3] dyes = MLJBase.UnivariateFinite([yes, no], [0.6, 0.4]) dno = MLJBase.UnivariateFinite([yes, no], [0.3, 0.7]) yhat = [dno, dno, dyes, dyes] w = [1, 2, 3, 4] @test MLJBase.ZeroOneLoss()(yhat, y) ≈ [1, 1, 1, 0] @test MLJBase.zero_one_loss(yhat,y, w) ≈ [1, 2, 3, 0] N = 10 y = categorical(rand(rng, ["yes", "no"], N), ordered=true) levels!(y, ["no", "yes"]) no, yes = MLJBase.classes(y[1]) @test pm1([yes, no]) in [[+1, -1], [-1, +1]] ym = pm1(y) # observations for raw LossFunctions measure p_vec = rand(N) yhat = MLJBase.UnivariateFinite([no, yes], p_vec, augment=true) yhatm = MLJBase._scale.(p_vec) # predictions for raw LossFunctions measure w = rand(rng, N) for M_ex in MARGIN_LOSSES m = eval(:(MLJBase.$M_ex())) @test m(yhat, y) ≈ LossFunctions.value(getfield(m, :loss), ym, yhatm) @test MLJBase.Mean()(m(yhat, y, w)) ≈ LossFunctions.value(getfield(m, :loss), ym, yhatm, WeightedSum(w))/N end end @testset "LossFunctions.jl - continuous" begin # losses for continuous targets: N = 10 y = randn(rng, N) yhat = randn(rng, N) X = nothing w = rand(rng, N) for M_ex in DISTANCE_LOSSES m = eval(:(MLJBase.$M_ex())) m_ex = MLJBase.snakecase(M_ex) @test m == eval(:(MLJBase.$m_ex)) @test m(yhat, y) ≈ LossFunctions.value(getfield(m, :loss), y, yhat) @test mean(m(yhat ,y, w)) ≈ LossFunctions.value(getfield(m, :loss), y, yhat, WeightedSum(w))/N end end
{ "alphanum_fraction": 0.5810260586, "author": null, "avg_line_length": 33.1891891892, "converted": null, "ext": "jl", "file": null, "hexsha": "30fe057e61ad0278776f921042bbc71acaa066f0", "include": null, "lang": "Julia", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "2e9444d64c6e1f63c0b2be409e1aadb7692672b7", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "Rahulub3r/MLJBase.jl", "max_forks_repo_path": "test/measures/loss_functions_interface.jl", "max_issues_count": null, "max_issues_repo_head_hexsha": "2e9444d64c6e1f63c0b2be409e1aadb7692672b7", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "Rahulub3r/MLJBase.jl", "max_issues_repo_path": "test/measures/loss_functions_interface.jl", "max_line_length": 78, "max_stars_count": null, "max_stars_repo_head_hexsha": "2e9444d64c6e1f63c0b2be409e1aadb7692672b7", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "Rahulub3r/MLJBase.jl", "max_stars_repo_path": "test/measures/loss_functions_interface.jl", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 828, "path": null, "reason": null, "repo": null, "save_path": null, "sha": null, "size": 2456 }
import numpy as np import pandas as pd data = pd.read_csv('input.txt', sep="\n", header=None) # Part 1 def calculate_fuel_required(mass): return int(mass / 3) - 2 # divide by 3, round down, subtract 2 fuel = data.apply(calculate_fuel_required, axis=1) fuel_required_sum = fuel.values.sum() print("Fuel required = {0}".format(fuel_required_sum)) # Part 2 def calculate_fuel_recursive(mass): fuel_required = calculate_fuel_required(mass) if fuel_required > 0: return fuel_required + calculate_fuel_recursive(fuel_required) return 0 recursive_fuel = data.apply(calculate_fuel_recursive, axis=1) recursive_fuel_sum = recursive_fuel.values.sum() print("Fuel required (recursive) = {0}".format(recursive_fuel_sum))
{ "alphanum_fraction": 0.7489878543, "author": null, "avg_line_length": 27.4444444444, "converted": null, "ext": "py", "file": null, "hexsha": "9ec4e921aeefccda9c37f646f2c75a516876e76d", "include": true, "lang": "Python", "length": null, "llama_tokens": null, "mathlib_filename": null, "max_forks_count": null, "max_forks_repo_forks_event_max_datetime": null, "max_forks_repo_forks_event_min_datetime": null, "max_forks_repo_head_hexsha": "28bd937d08dafbd232e4818d11cb37767a91d9c6", "max_forks_repo_licenses": [ "MIT" ], "max_forks_repo_name": "timothyr/advent_of_code_2019", "max_forks_repo_path": "day1/solution.py", "max_issues_count": null, "max_issues_repo_head_hexsha": "28bd937d08dafbd232e4818d11cb37767a91d9c6", "max_issues_repo_issues_event_max_datetime": null, "max_issues_repo_issues_event_min_datetime": null, "max_issues_repo_licenses": [ "MIT" ], "max_issues_repo_name": "timothyr/advent_of_code_2019", "max_issues_repo_path": "day1/solution.py", "max_line_length": 70, "max_stars_count": null, "max_stars_repo_head_hexsha": "28bd937d08dafbd232e4818d11cb37767a91d9c6", "max_stars_repo_licenses": [ "MIT" ], "max_stars_repo_name": "timothyr/advent_of_code_2019", "max_stars_repo_path": "day1/solution.py", "max_stars_repo_stars_event_max_datetime": null, "max_stars_repo_stars_event_min_datetime": null, "num_tokens": 195, "path": null, "reason": "import numpy", "repo": null, "save_path": null, "sha": null, "size": 741 }