id
stringlengths
3
8
content
stringlengths
100
981k
11491504
import request_generator class MouseRequest(request_generator.RequestGenerator): def __init__(self, client): super(MouseRequest, self).__init__(client) self.type = 'action' self.device = 'mouse' def press(self, mask): self.action = 'press' self.params = [mask] self.send_request() def release(self, mask): self.action = 'release' self.params = [mask] self.send_request() def click(self, mask): self.press(mask) self.release(mask) def left_click(self, x = None, y = None): self.action = 'left_lick' if x is None or y is None: self.params = [] else: self.params = [x, y] return self.send_request() def right_click(self, x = None, y = None): self.action = 'right_click' if x is None or y is None: self.params = [] else: self.params = [x, y] return self.send_request() def move(self, x, y): self.action = 'move' self.params = [x,y] return self.send_request() def move_by(self, x, y): self.action = 'move_by' self.params = [x,y] return self.send_request() def drag(self, x1, y1, x2, y2): self.action = 'drag' self.params = [x1, y1, x2, y2] return self.send_request() def drag_by(self, x, y): self.action = 'drag_by' self.params = [x,y] return self.send_request() def get_position(self): self.action = 'get_position' self.params = [] return self.send_request() def get_color(self, x = None, y = None): self.action = 'get_color' if x is None or y is None: self.params = [] else: self.params = [x, y] return self.send_request()
11491527
import aiohttp import asyncio from typing import Awaitable, Callable from tests.benchmark.driver_base import * TestTask = Callable[[aiohttp.ClientSession], Awaitable[int]] async def run_notify_test(session: aiohttp.ClientSession) -> int: async with session.post(notify_url(), json=notify_entity()) as response: return response.status async def run_version_test(session: aiohttp.ClientSession) -> int: async with session.get(version_url()) as response: return response.status # NOTE. Timing coroutines. # Starting a timer before `async with` and then stopping it in the block # won't measure what you think. E.g. # # sample_id = monitor.start_duration_sample() # async with session.get(version_url()) as response: # label = f"client:version:{response.status}" # monitor.stop_duration_sample(label, sample_id) # # won't actually time just how long the HTTP request took from start to # finish, but will also include the time the various coroutines sat waiting # in the event loop. While there's no accurate way of timing coroutines that # I know, in the specific case of aiohttp, we could provide some half-meaningful # measurements: # - https://stackoverflow.com/questions/46004745 def lookup_test_task(test_id: str) -> TestTask: tasks = { VERSION_TEST: run_version_test, NOTIFY_TEST: run_notify_test } return tasks[test_id] async def do_many(task: TestTask, how_many: int) -> TestRunResults: async with aiohttp.ClientSession() as session: tasks = [task(session) for _ in range(how_many)] return await asyncio.gather(*tasks, return_exceptions=True) class AsyncioDriver(Driver): def _do_run(self, test_id: str) -> TestRunResults: test_task = lookup_test_task(test_id) return asyncio.run(do_many(test_task, REQUESTS_N)) if __name__ == "__main__": AsyncioDriver().main()
11491551
from amazon.api import AmazonAPI from ebaysdk.finding import Connection as finding from ebaysdk.exception import ConnectionError import csv from fuzzywuzzy import fuzz import argparse parser = argparse.ArgumentParser(description='Description of your program') parser.add_argument('-S','--SearchTerm', help='what you would like to search for', required=True) args = vars(parser.parse_args()) #keys AWSkey = 'your AWS key here' AWSSecretKey = 'your AWS secret key here' amazonProductAdvName = 'your amazon product advertising name here' EbayKey = 'your ebay developer key here' search = u'%s'%args['SearchTerm'] amazon = AmazonAPI(AWSkey,AWSSecretKey,amazonProductAdvName,MaxQPS=.9) ebayApi = finding(appid=EbayKey, config_file=None) path = open("AmazonEbayData.csv", "wt") filewriter = csv.writer(path) allData = [] filewriter.writerow(['Ebay Item Title', 'Ebay Item ID','Ebay Price','ebay Item URL', 'Amazon Item Title','Amazon ISBN', 'Amazon Price', 'Amazon Item URL', 'Price Difference']) def findOnAmazon(ebayItem,ebayPrice,ebayItemID,ebayItemURL): response = amazon.search_n(1,Keywords=ebayItem, SearchIndex='All') amazonPrice = response[0].price_and_currency[0] amazonTitle = response[0].title amazonISBN = response[0].isbn amazonOfferURL = response[0].offer_url #check how close the titles are so we can better understand if they are the same product r = fuzz.partial_ratio(ebayItem,amazonTitle) if r > 75: # we can assume that they are the same thing if float(amazonPrice) > float(ebayPrice)*1.5: # check if the item is at least 50% more expensive on amazon so we have room to make money possibleReturn = float(amazonPrice) - float(ebayPrice) possret = float("{0:.2f}".format(possibleReturn)) print(possret) data = [ebayItem,ebayItemID,ebayPrice,ebayItemURL,amazonTitle,amazonISBN,amazonPrice,amazonOfferURL,possret] allData.append(data) def run(): try: api_request = { 'keywords': search, 'itemFilter': [ {'name': 'LocatedIn', 'value': 'US'}, {'name': 'MinPrice', 'value': '0'}, {'name': 'MaxPrice', 'value': '50.00'}, ], 'sortOrder': 'BestMatch', } response = ebayApi.execute('findItemsAdvanced', api_request) dict = response.reply.get('searchResult') clearedArray = dict.get('item') for item in clearedArray: ebaytitle = item.get('title') ebayItemID = item.get('itemId') ebayprice = item.get('sellingStatus') current = ebayprice.get('currentPrice') value = current.get('value') ebayURL = item.get('viewItemURL') try: findOnAmazon(ebaytitle,value,ebayItemID,ebayURL) except: print("couldn`t find matching object") for row in allData: filewriter.writerow(row) path.close() except ConnectionError as e: print(e) print(e.response.dict()) if __name__ == '__main__': if args['SearchTerm']: run()
11491597
import functools import torch from scipy.linalg import lapack as scll from falkon.la_helpers import potrf from falkon.options import FalkonOptions from falkon.utils.helpers import choose_fn __all__ = ("check_init", "inplace_set_diag_th", "inplace_add_diag_th", "lauum_wrapper", "potrf_wrapper") def check_init(*none_check): def _checker(fun): @functools.wraps(fun) def wrapper(self, *args, **kwargs): is_init = True for el in none_check: if getattr(self, el, None) is None: is_init = False break if not is_init: raise RuntimeError( "FALKON preconditioner is not initialized. Please run " "`init` before any other method on the " "preconditioner.") return fun(self, *args, **kwargs) return wrapper return _checker def inplace_set_diag_th(A: torch.Tensor, k: torch.Tensor) -> torch.Tensor: A.diagonal().copy_(k) return A def inplace_add_diag_th(A: torch.Tensor, k: float) -> torch.Tensor: # Assumes M is square (or wide also works). # Need to use .diagonal() as .diag() makes a copy A.diagonal().add_(k) return A def lauum_wrapper(A: torch.Tensor, upper: bool, use_cuda: bool, opt: FalkonOptions) -> torch.Tensor: if use_cuda: from falkon.ooc_ops.ooc_lauum import gpu_lauum return gpu_lauum(A, upper=upper, write_opposite=True, overwrite=True, opt=opt) else: Anp = A.numpy() lauum = choose_fn(Anp.dtype, scll.dlauum, scll.slauum, "LAUUM") sol, info = lauum(Anp, lower=int(not upper), overwrite_c=1) if info != 0: raise RuntimeError(f"Lapack LAUUM failed with error code {info}.") return torch.from_numpy(sol) def potrf_wrapper(A: torch.Tensor, clean: bool, upper: bool, use_cuda: bool, opt: FalkonOptions) -> torch.Tensor: if use_cuda: from falkon.ooc_ops.ooc_potrf import gpu_cholesky return gpu_cholesky(A, upper=upper, clean=clean, overwrite=True, opt=opt) else: return potrf(A, upper=upper, clean=clean, overwrite=True, cuda=False)
11491620
import config import subprocess from flask import request from flask_restful import Resource, reqparse import logging import base64 from decorators import private_api from requests import get import json import shlex logger = logging.getLogger("api") class Jobs(Resource): @private_api def get(self): """ List all jobs in the queue --- tags: - Scheduler responses: 200: description: List of all jobs 500: description: Backend error """ parser = reqparse.RequestParser() parser.add_argument('user', type=str, location='args') args = parser.parse_args() user = args['user'] try: qstat_command = config.Config.PBS_QSTAT + " -f -Fjson" try: get_job_info = subprocess.check_output(shlex.split(qstat_command)) try: job_info = json.loads(((get_job_info.decode('utf-8')).rstrip().lstrip())) except Exception as err: return {"success": False, "message": "Unable to retrieve this job. Job may have terminated."}, 500 if user is None: return {"success": True, "message": job_info["Jobs"] if "Jobs" in job_info.keys() else {}}, 200 else: job_for_user = {"Jobs": {}} if "Jobs" in job_info.keys(): job_ids_key = list(job_info["Jobs"].keys()) for job_id in job_ids_key: job_owner = job_info["Jobs"][job_id]["Job_Owner"].split("@")[0] if job_owner == user: job_for_user["Jobs"][job_id] = job_info["Jobs"][job_id] return {"success": True, "message": job_for_user["Jobs"]}, 200 except Exception as err: return {"succes": False, "message": "Unable to retrieve Job ID (job may have terminated and is no longer in the queue)"}, 500 except Exception as err: return {"success": False, "message": "Unknown error: " + str(err)}, 500
11491621
from .c11type import C11Type class C11TypeBool(C11Type): def __init__(self): C11Type.__init__(self) self.typeName = u'bool' def setSchema(self, schemaName, schemaValue): C11Type.setSchema(self, schemaName, schemaValue) @classmethod def codeDefaultValue(cls, schemaDefaultValue): if schemaDefaultValue is not None and schemaDefaultValue is True: return u'true' return u'false' @classmethod def codeDefaultValueArray(cls, schemaDefaultValues): if schemaDefaultValues is None\ or not isinstance(schemaDefaultValues, list)\ or len(schemaDefaultValues) <= 0: return u'' code_default_value = u'' for schema_value in schemaDefaultValues: if len(code_default_value) > 0: code_default_value = code_default_value + u', ' if schema_value == True: code_default_value = code_default_value + u'true' else: code_default_value = code_default_value + u'false' return u'{ %s }' % code_default_value @classmethod def codeJsonCheck(cls): return u'IsBool()' @classmethod def codeJsonSet(cls, dataName, variableName): return u'%s.%s = _JsonValue[GLTFTEXT("%s")].GetBool();' % (dataName, variableName, variableName) @classmethod def codeJsonGet(cls, dataName, variableName): return u'_JsonValue[GLTFTEXT("%s")].SetBool(%s.%s);' % (variableName, dataName, variableName)
11491622
import setuptools pkg_name="theiapod" setuptools.setup( name=pkg_name, version="0.1.0", author="<NAME>", author_email="<EMAIL>", description="Self-host gitpod-style workspaces for github repositories using theia", url="https://github.com/magland/theiapod", packages=setuptools.find_packages(), scripts=['bin/theiapod'], package_data={}, install_requires=[ "pyyaml", ], classifiers=( "Programming Language :: Python :: 3", "License :: OSI Approved :: Apache Software License", "Operating System :: OS Independent", ) )
11491635
import pytest import os from tests.examples import TestExample as base_class from pandas.testing import assert_frame_equal from yggdrasil.components import create_component class TestExampleConditionalIO(base_class): r"""Test the conditional_io example.""" @pytest.fixture(scope="class") def example_name(self): r"""str: Name of example being tested.""" return "conditional_io" @pytest.fixture def expected_output_files(self, yamldir): r"""list: Examples of expected output for the run.""" return [os.path.join(yamldir, 'Output', 'output.txt')] @pytest.fixture def output_files(self, yamldir): r"""list: Output files for the run.""" return [os.path.join(yamldir, 'output.txt')] @pytest.fixture(scope="class") def read_file(self): r"""Read in contents from a file. Args: fname (str): Full path to the file that should be read. Returns: object: File contents. """ def read_file_w(fname): x = create_component('file', 'table', name='test', address=fname, direction='recv', as_array=True, recv_converter='pandas') msg = x.recv_array()[1] if msg is not None: msg = msg.sort_values(by=['InputMass']).reset_index( drop=True) x.close() return msg return read_file_w @pytest.fixture(scope="class") def check_file_contents(self, read_file): r"""Check that the contents of a file are correct. Args: fname (str): Full path to the file that should be checked. result (str): Contents of the file. """ def check_file_contents_w(fname, result): ocont = read_file(fname) assert_frame_equal(ocont, result) return check_file_contents_w @pytest.fixture(scope="class") def check_file_size(self): r"""Check that file is the correct size. Args: fname (str): Full path to the file that should be checked. fsize (int): Size that the file should be in bytes. """ def check_file_size_w(*args, **kwargs): pass return check_file_size_w
11491639
import os from tqdm import trange import torch from im2mesh.common import chamfer_distance from im2mesh.training import BaseTrainer from im2mesh.utils import visualize as vis class Trainer(BaseTrainer): r''' Trainer object for the Point Set Generation Network. The PSGN network is trained on Chamfer distance. The Trainer object obtains methods to perform a train and eval step as well as to visualize the current training state by plotting the respective point clouds. Args: model (nn.Module): PSGN model optiimzer (PyTorch optimizer): The optimizer that should be used device (PyTorch device): the PyTorch device input_type (string): The input type (e.g. 'img') vis_dir (string): the visualisation directory ''' def __init__(self, model, optimizer, device=None, input_type='img', vis_dir=None): self.model = model self.optimizer = optimizer self.device = device self.input_type = input_type self.vis_dir = vis_dir if vis_dir is not None and not os.path.exists(vis_dir): os.makedirs(vis_dir) def train_step(self, data): r''' Performs a train step. The chamfer loss is calculated and an appropriate backward pass is performed. Args: data (tensor): training data ''' self.model.train() points = data.get('pointcloud').to(self.device) inputs = data.get('inputs').to(self.device) loss = self.compute_loss(points, inputs) self.optimizer.zero_grad() loss.backward() self.optimizer.step() return loss.item() def eval_step(self, data): r''' Performs an evaluation step. The chamfer loss is calculated and returned in a dictionary. Args: data (tensor): input data ''' self.model.eval() device = self.device points = data.get('pointcloud_chamfer').to(device) inputs = data.get('inputs').to(device) with torch.no_grad(): points_out = self.model(inputs) loss = chamfer_distance(points, points_out).mean() loss = loss.item() eval_dict = { 'loss': loss, 'chamfer': loss, } return eval_dict def visualize(self, data): r''' Visualizes the current output data of the model. The point clouds for respective input data is plotted. Args: data (tensor): input data ''' device = self.device points_gt = data.get('pointcloud').to(device) inputs = data.get('inputs').to(device) with torch.no_grad(): points_out = self.model(inputs) points_out = points_out.cpu().numpy() points_gt = points_gt.cpu().numpy() batch_size = inputs.size(0) for i in trange(batch_size): input_img_path = os.path.join(self.vis_dir, '%03d_in.png' % i) vis.visualize_data( inputs[i].cpu(), self.input_type, input_img_path) out_file = os.path.join(self.vis_dir, '%03d.png' % i) out_file_gt = os.path.join(self.vis_dir, '%03d_gt.png' % i) vis.visualize_pointcloud(points_out[i], out_file=out_file) vis.visualize_pointcloud(points_gt[i], out_file=out_file_gt) def compute_loss(self, points, inputs): r''' Computes the loss. The Point Set Generation Network is trained on the Chamfer distance. Args: points (tensor): GT point cloud data inputs (tensor): input data for the model ''' points_out = self.model(inputs) loss = chamfer_distance(points, points_out).mean() return loss
11491681
from plugin.core.constants import GUID_SERVICES import logging log = logging.getLogger(__name__) class SyncMap(object): def __init__(self, task): self.task = task self._by_guid = {} self._by_key = {} def add(self, p_section_key, p_key, guids): if not guids or not p_key: return False for guid in guids: self.add_one(p_section_key, p_key, guid) def add_one(self, p_section_key, p_key, guid): if guid is None or p_key is None: return False p_key = int(p_key) # Flatten `guid` if type(guid) is not tuple: guid = (guid.service, guid.id) if guid[0] not in GUID_SERVICES: log.info('Unknown primary agent: %r -> %r (section: %r)', guid[0], p_key, p_section_key) # Store in `_by_guid` map if guid not in self._by_guid: self._by_guid[guid] = set() self._by_guid[guid].add((p_section_key, p_key)) # Store in `_by_key` map if p_key not in self._by_key: self._by_key[p_key] = set() self._by_key[p_key].add(guid) return True def by_guid(self, guid): # Flatten `guid` if type(guid) is not tuple: guid = (guid.service, guid.id) return self._by_guid.get(guid, set()) def by_key(self, rating_key): if rating_key is None: return set() return self._by_key.get(int(rating_key), set())
11491702
import unittest import os import_error = False try: from .. import common except ImportError: import_error = True common = None class TestCase00(unittest.TestCase): def test_import(self): self.assertFalse(import_error) class TestCase01(unittest.TestCase): def setUp(self): if import_error: self.skipTest('ImportError') def test_backup1(self): path = common.backup('common.pz', copy=False) self.assertEqual(path, '') def test_backup2(self): path = common.backup(os.path.join(os.path.dirname(__file__), 'test_common.py'), copy=False) self.assertEqual(os.path.dirname(path), os.path.dirname(__file__)) self.assertEqual(os.path.basename(path), 'test_common.py.1')
11491734
import FWCore.ParameterSet.Config as cms from DQMOffline.JetMET.metDQMConfig_cfi import * #correction for type 1 done in JetMETDQMOfflineSource now METDQMAnalyzerSequence = cms.Sequence(caloMetDQMAnalyzer*pfMetDQMAnalyzer*pfChMetDQMAnalyzer*pfMetT1DQMAnalyzer) METDQMAnalyzerSequenceMiniAOD = cms.Sequence(pfMetDQMAnalyzerMiniAOD*pfPuppiMetDQMAnalyzerMiniAOD) METDQMAnalyzerSequenceCosmics = cms.Sequence(caloMetDQMAnalyzer) METDQMAnalyzerSequenceHI = cms.Sequence(caloMetDQMAnalyzer*pfMetDQMAnalyzer)
11491737
from __future__ import (absolute_import, division, print_function) __metaclass__ = type import yaml from ansible.plugins.action import ActionBase from ansible.errors import AnsibleActionFail from ansible.utils.vars import isidentifier from ansible.plugins.filter.core import combine from ansible.plugins.loader import lookup_loader from ansible_collections.arista.avd.plugins.module_utils.strip_empties import strip_null_from_data class ActionModule(ActionBase): def run(self, tmp=None, task_vars=None): if task_vars is None: task_vars = {} result = super().run(tmp, task_vars) del tmp # tmp no longer has any effect root_key = "" if self._task.args: if "root_key" in self._task.args: n = self._task.args.get("root_key") n = self._templar.template(n) if not isidentifier(n): raise AnsibleActionFail(f"The argument 'root_key' value of '{n}' is not valid. Keys must start with a letter or underscore character, " "and contain only letters, numbers and underscores.") root_key = n if "templates" in self._task.args: t = self._task.args.get("templates") if isinstance(t, list): template_list = t else: raise AnsibleActionFail("The argument 'templates' is not a list") else: raise AnsibleActionFail("The argument 'templates' must be set") else: raise AnsibleActionFail("The argument 'templates' must be set") output = {} template_lookup_module = lookup_loader.get('ansible.builtin.template', loader=self._loader, templar=self._templar) template_vars = task_vars for template_item in template_list: template = template_item.get('template') if not template: raise AnsibleActionFail("Invalid template data") template_options = template_item.get('options', {}) list_merge = template_options.get('list_merge', 'append') strip_empty_keys = template_options.get('strip_empty_keys', True) if root_key: template_vars[root_key] = output else: template_vars = combine(task_vars, output, recursive=True) template_output = template_lookup_module.run([template], template_vars) template_output_data = yaml.safe_load(template_output[0]) if strip_empty_keys: template_output_data = strip_null_from_data(template_output_data) if template_output_data: output = combine(output, template_output_data, recursive=True, list_merge=list_merge) if root_key: result['ansible_facts'] = {root_key: output} else: result['ansible_facts'] = output return result
11491760
import dash import dash_core_components as dcc import dash_html_components as html from dash.dependencies import Input, Output from textwrap import dedent as d import pandas as pd #df = pd.read_hdf(r'..\examples\algae\MC\mc_out_allyears_sigma1.h5') df = pd.read_csv(r'..\examples\algae\MC\mc_out_allyears_sigma1.dat', delim_whitespace=True) data = df.values param = data[:,0:8] def generate_table(): data = dict() for l in range(4): for k in range(2): data[k+2*l] = [ { 'x': df.iloc[::100, k+2*l], 'type': 'histogram' } ] return html.Table( [html.Tr([ html.Td(dcc.Graph(id='hist{}'.format(k+2*l), figure = {'data': data[k+2*l], 'layout': {}}), style={'width': '500px'}) for k in range(2) ], style={'height': '300px'}) for l in range(4)], ) external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css'] app = dash.Dash(__name__, external_stylesheets=external_stylesheets) app.layout = html.Div(children=[ html.H1(children='Conditional probability distributions'), html.Div([ dcc.Markdown(d(""" **Click** on bars in one histogram to plot conditional probability distribution for other variables. """)), html.Button('Reset', id='reset', n_clicks_timestamp=0), #html.Pre(id='click-data'), ]), generate_table() ]) #@<EMAIL>.callback( # Output('click-data', 'children'), # [Input('hist0', 'clickData')]) #def display_click_data(clickData): #return json.dumps(clickData['points'][0], indent=2) # try: # return json.dumps(clickData['points'][0]['binNumber']) # except: # return 0 def gen_callback(k): from numpy import array def update_figure(*args): defaultdata = { 'data': [ { 'x': df.iloc[::100, k], 'type': 'histogram' } ] } # Find out who triggered the callback, # see https://github.com/plotly/dash/issues/291 ctx = dash.callback_context # Update was not triggered at all if not ctx.triggered: return defaultdata trigger = ctx.triggered[0] # Reset button triggered update if trigger['prop_id'] == 'reset.n_clicks': return defaultdata # Other component triggered update try: reducedset = array(trigger['value']['points'][0]['pointNumbers']) except: return defaultdata return { 'data': [ { 'x': df.iloc[reducedset, k], 'type': 'histogram' } ] } return update_figure for k in range(8): app.callback( output = Output('hist{}'.format(k), 'figure'), inputs = ( [Input('hist{}'.format(l), 'clickData') for l in range(8)] + [Input('reset', 'n_clicks')]))(gen_callback(k)) if __name__ == '__main__': app.run_server(debug=True, host='0.0.0.0')
11491761
import numpy as np import matplotlib.pyplot as plt from pactools.utils.testing import assert_equal, assert_array_almost_equal from pactools.utils.testing import assert_greater, assert_raises from pactools.utils.testing import assert_array_not_almost_equal from pactools.dar_model import DAR, AR, HAR, StableDAR from pactools.simulate_pac import simulate_pac ALL_MODELS = [DAR, AR, HAR, StableDAR] # Parameters used for the simulated sigin in the test low_fq_range = [1., 3., 5., 7.] high_fq_range = [25., 50., 75.] n_low = len(low_fq_range) n_high = len(high_fq_range) high_fq = high_fq_range[1] low_fq = low_fq_range[1] n_points = 1024 fs = 200. _sigin = simulate_pac(n_points=n_points, fs=fs, high_fq=high_fq, low_fq=low_fq, low_fq_width=1., noise_level=0.3, random_state=0) _sigdriv = simulate_pac(n_points=n_points, fs=fs, high_fq=high_fq, low_fq=low_fq, low_fq_width=1., noise_level=0., random_state=0, high_fq_amp=0, return_driver=True) _sigdriv_imag = np.imag(_sigdriv) _sigdriv = np.real(_sigdriv) _noise = np.random.RandomState(0).randn(n_points) _model_params = {'ordar': 10, 'ordriv': 2, 'criterion': False} def fast_fitted_model(klass=DAR, model_params=_model_params, sigin=_sigin, sigdriv=_sigdriv, sigdriv_imag=_sigdriv_imag, fs=fs, train_weights=None, test_weights=None): model_params = model_params.copy() if klass == StableDAR and 'iter_newton' not in model_params: model_params['iter_newton'] = 10 return klass(**model_params).fit( sigin=sigin, sigdriv=sigdriv, sigdriv_imag=sigdriv_imag, fs=fs, train_weights=train_weights, test_weights=test_weights) def test_likelihood_ratio_noise(): # Test that likelihood_ratio returns a positive p_value with noise sdar = fast_fitted_model(StableDAR, sigin=_noise) dar = fast_fitted_model(DAR, sigin=_noise) ar = fast_fitted_model(AR, sigin=_noise) har = fast_fitted_model(HAR, sigin=_noise) for h0 in [ar, har]: for h1 in [dar, sdar]: p_value = h1.likelihood_ratio(h0) assert_greater(p_value, 0) def test_likelihood_ratio_pac(): # Test that likelihood_ratio returns a near zero p_value with PAC sdar = fast_fitted_model(StableDAR, sigin=_sigin) dar = fast_fitted_model(DAR, sigin=_sigin) ar = fast_fitted_model(AR, sigin=_sigin) har = fast_fitted_model(HAR, sigin=_sigin) for h0 in [ar, har]: for h1 in [dar, sdar]: p_value = h1.likelihood_ratio(h0) assert_greater(1e-3, p_value) def test_degrees_of_freedom(): # test the number of fitted parameters in the model model_params = {'ordar': 5, 'ordriv': 2, 'criterion': False} dar = fast_fitted_model(model_params=model_params) assert_equal(dar.degrees_of_freedom(), 6 * 6) model_params = {'ordar': 5, 'ordriv': 0, 'criterion': False} dar = fast_fitted_model(model_params=model_params) assert_equal(dar.degrees_of_freedom(), 6 * 1) model_params = {'ordar': 1, 'ordriv': 2, 'criterion': False} dar = fast_fitted_model(model_params=model_params) assert_equal(dar.degrees_of_freedom(), 2 * 6) model_params = {'ordar': 10, 'ordriv': 2, 'criterion': False} ar = fast_fitted_model(AR, model_params=model_params) assert_equal(ar.degrees_of_freedom(), 10 + 1) model_params = {'ordar': 10, 'ordriv': 2, 'criterion': False} har = fast_fitted_model(HAR, model_params=model_params) assert_equal(har.degrees_of_freedom(), 10 + 6) model_params = {'ordar': 5, 'ordriv': 2, 'criterion': False} har = fast_fitted_model(model_params=model_params, sigdriv_imag=None) assert_equal(har.degrees_of_freedom(), 6 * 3) model_params = {'ordar': 10, 'ordriv': 2, 'criterion': False} har = fast_fitted_model(HAR, model_params=model_params, sigdriv_imag=None) assert_equal(har.degrees_of_freedom(), 10 + 3) def dar_no_fit(ortho, normalize, sigdriv=_sigdriv, sigdriv_imag=_sigdriv_imag, **model_params): dar = DAR(ortho=ortho, normalize=normalize, **model_params) dar.sigin = _sigin[None, :] dar.sigdriv = sigdriv[None, :] dar.sigdriv_imag = sigdriv_imag[None, :] dar._make_basis() return dar def test_make_basis_new_sigdriv(): # Test that _make_basis works the same with a new sigdriv, # using stored orthonormalization transform. model_params = {'ordar': 5, 'ordriv': 2, 'criterion': False} for normalize in (True, False): for ortho in (True, False): for this_sigdriv, this_sigdriv_imag in ([_sigdriv, _sigdriv_imag], [_noise, _noise[::-1]]): dar = dar_no_fit( ortho=ortho, normalize=normalize, sigdriv=this_sigdriv, sigdriv_imag=this_sigdriv_imag, **model_params) newbasis = dar._make_basis(sigdriv=this_sigdriv, sigdriv_imag=this_sigdriv_imag) assert_array_almost_equal(newbasis, dar.basis_) # Different result if we change a parameter dar_ortho = dar_no_fit(not ortho, normalize, **model_params) dar_norma = dar_no_fit(ortho, not normalize, **model_params) for dar2 in [dar_ortho, dar_norma]: assert_raises(AssertionError, assert_array_almost_equal, dar.basis_, dar2.basis_) def test_make_basis_ortho_normalize(): # Test the effect of ortho and normalize parameters model_params = {'ordar': 5, 'ordriv': 2, 'criterion': False} for normalize in (True, False): for ortho in (True, False): dar = dar_no_fit(ortho, normalize, **model_params) basis = dar.basis_ basis.shape = (basis.shape[0], -1) product = np.dot(basis, basis.T) n_basis = product.shape[0] n_samples = _sigin.size # Test that the diagonal is constant if normalized ref = np.ones(n_basis) * n_samples if normalize: assert_array_almost_equal(product.flat[::n_basis + 1], ref) else: assert_array_not_almost_equal(product.flat[::n_basis + 1], ref) # Test that the rest is zero if orthogonalized product.flat[::n_basis + 1] = 0 ref = np.zeros((n_basis, n_basis)) if ortho: assert_array_almost_equal(product, ref) else: assert_array_not_almost_equal(product, ref) def test_plot_comodulogram(): # Smoke test with the standard plotting function model = fast_fitted_model() model.plot_lines() model.plot() plt.close('all') model = fast_fitted_model(sigdriv_imag=None) model.plot_lines() model.plot() plt.close('all') def test_weighting_with_ones(): # Test that weighting with ones is identical as no weighting factor = 1.5 train_weights = np.ones_like(_sigin) * factor for klass in ALL_MODELS: model_2 = fast_fitted_model(klass=klass, train_weights=train_weights, test_weights=train_weights) model_1 = fast_fitted_model(klass=klass, train_weights=train_weights, test_weights=None) model_0 = fast_fitted_model(klass=klass, train_weights=None, test_weights=None) assert_array_almost_equal(model_1.AR_, model_0.AR_, decimal=5) assert_array_almost_equal(model_2.AR_, model_0.AR_, decimal=5) assert_array_almost_equal(model_1.G_, model_0.G_, decimal=5) assert_array_almost_equal(model_2.G_, model_0.G_, decimal=5) for train in [False, True]: assert_array_almost_equal( model_0._estimate_log_likelihood(train=train)[0] * factor, model_1._estimate_log_likelihood(train=train)[0], decimal=4) assert_array_almost_equal( model_1._estimate_log_likelihood(train=train)[0], model_2._estimate_log_likelihood(train=train)[0], decimal=4) def test_weighting_with_zeros(): # Test that split = int(n_points // 2) train_weights = np.ones_like(_sigin) train_weights[split:] = 0 sigin_half = _sigin[:split] sigdriv_half = _sigdriv[:split] sigdriv_imag_half = _sigdriv_imag[:split] for klass in ALL_MODELS: print(klass.__name__) model_1 = fast_fitted_model(klass=klass, train_weights=train_weights) model_0 = fast_fitted_model( klass=klass, sigin=sigin_half, sigdriv=sigdriv_half, sigdriv_imag=sigdriv_imag_half, train_weights=None) assert_array_almost_equal(model_1.AR_, model_0.AR_) assert_array_almost_equal(model_1.G_, model_0.G_) for train in [False, True]: assert_array_almost_equal( model_0._estimate_log_likelihood(train=train), model_1._estimate_log_likelihood(train=train), decimal=5)
11491778
import face_alignment import numpy as np import cv2 import threading import cv2 class VideoCaptureThreading: def __init__(self, src=0, width=640, height=480): self.src = src self.cap = cv2.VideoCapture(self.src) self.cap.set(cv2.CAP_PROP_FRAME_WIDTH, width) self.cap.set(cv2.CAP_PROP_FRAME_HEIGHT, height) self.grabbed, self.frame = self.cap.read() self.started = False self.read_lock = threading.Lock() def set(self, var1, var2): self.cap.set(var1, var2) def start(self): if self.started: print("[!] Threaded video capturing has already been started.") return None self.started = True self.thread = threading.Thread(target=self.update, args=()) self.thread.start() return self def update(self): while self.started: grabbed, frame = self.cap.read() with self.read_lock: self.grabbed = grabbed self.frame = frame def read(self): with self.read_lock: frame = self.frame.copy() grabbed = self.grabbed return grabbed, frame def stop(self): self.started = False self.thread.join() def __exit__(self, exec_type, exc_value, traceback): self.cap.release() def bb_intersection_over_union(boxA, boxB): xA = max(boxA[0], boxB[0]) yA = max(boxA[1], boxB[1]) xB = min(boxA[2], boxB[2]) yB = min(boxA[3], boxB[3]) interArea = max(0, xB - xA + 1) * max(0, yB - yA + 1) boxAArea = (boxA[2] - boxA[0] + 1) * (boxA[3] - boxA[1] + 1) boxBArea = (boxB[2] - boxB[0] + 1) * (boxB[3] - boxB[1] + 1) iou = interArea / float(boxAArea + boxBArea - interArea) return iou def extract_bbox(frame, refbbox, fa): bboxes = fa.face_detector.detect_from_image(frame[..., ::-1]) if len(bboxes) != 0: bbox = max( [ (bb_intersection_over_union(bbox, refbbox), tuple(bbox)) for bbox in bboxes ] )[1] else: bbox = np.array([0, 0, 0, 0, 0]) return np.maximum(np.array(bbox), 0) def get_max_length(bbox): x1, y1, x2, y2 = bbox w = x2 - x1 h = y2 - y1 square_length = max(w, h) return square_length def get_face_bbox(image, face_detector): bboxes = face_detector.face_detector.detect_from_image(frame) if not bboxes: return [] growth_factor = 0.4 img_h, img_w = image.shape[:2] bbox = list(map(int, bboxes[0])) x1, y1, x2, y2, score = bbox square_length = get_max_length((x1, y1, x2, y2)) dlength = int(growth_factor * square_length) x1 = max(0, x1 - dlength) y1 = max(0, y1 - dlength) x2 = min(img_w - 1, x2 + dlength) y2 = min(img_h - 1, y2 + int(dlength * 0.5)) output_length = get_max_length((x1, y1, x2, y2)) return (x1, y1, x1 + output_length, y1 + output_length) fa = face_alignment.FaceAlignment( face_alignment.LandmarksType._2D, flip_input=True, device="cpu" ) # cap = cv2.VideoCapture(0) cap = VideoCaptureThreading(0) cap.start() bbox = [] # Capture initial video while True: grabbed, frame = cap.read() if not grabbed: break frame = frame[..., ::-1].copy() # bboxes = fa.face_detector.detect_from_image(frame) # print(bboxes) bbox = get_face_bbox(frame, fa) x1, y1, x2, y2 = bbox if bbox: # bbox = list(map(int, bboxes[0])) cv2.rectangle(frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]), (255, 255, 255)) cv2.imshow("original", frame[..., ::-1]) cv2.imshow("cropped", frame[y1:y2, x1:x2][..., ::-1]) key = cv2.waitKey(1) if key == 27: # ESC break # cap.release() cap.stop() cv2.destroyAllWindows()
11491792
import numpy as np from scipy.stats import mode, itemfreq from scipy import delete import matplotlib.pylab as plt from sklearn.linear_model import LogisticRegression from sklearn.ensemble import RandomForestClassifier from sklearn.svm import LinearSVC as SVM from missing_data_imputation import Imputer # declare csv headers x = np.genfromtxt('data/adult-train-raw', delimiter=', ', dtype=object) # remove redundant education-number feature x = delete(x, (4, 14), 1) # enumerate parameters and instantiate Imputer imp = Imputer() missing_data_cond = lambda x: x == '?' cat_cols = (1, 3, 4, 5, 6, 7, 8, 12) n_neighbors = 5 # # drop observations with missing variables # print 'imputing with drop' # data_drop = imp.drop(x, missing_data_cond) # replace missing values with random existing values print 'imputing with random replacement' data_replace = imp.replace(x, missing_data_cond) # replace missing values with feature summary print 'imputing with feature summarization (mode)' summ_func = lambda x: mode(x)[0] data_mode = imp.summarize(x, summ_func, missing_data_cond) # replace categorical features with one hot row print 'imputing with one-hot' data_onehot = imp.binarize_data(x, cat_cols) # replace missing data with predictions using random forest print 'imputing with predicted values from random forest' clf = RandomForestClassifier(n_estimators=100, criterion='gini') data_rf = imp.predict(x, cat_cols, missing_data_cond, clf) # replace missing data with predictions using SVM print 'imputing with predicted values usng SVM' clf = SVM( penalty='l2', loss='squared_hinge', dual=True, tol=0.0001, C=1.0, multi_class='ovr', fit_intercept=True, intercept_scaling=1, class_weight=None, verbose=0, random_state=None, max_iter=1000) data_svm = imp.predict(x, cat_cols, missing_data_cond, clf) # replace missing data with predictions using logistic regression print 'imputing with predicted values usng logistic regression' clf = LogisticRegression( penalty='l2', dual=False, tol=0.0001, C=1.0, fit_intercept=True, intercept_scaling=1) data_logistic = imp.predict(x, cat_cols, missing_data_cond, clf) # replace missing data with values obtained after factor analysis print 'imputing with factor analysis' data_facanal = imp.factor_analysis(x, cat_cols, missing_data_cond) # replace missing data with knn print 'imputing with K-Nearest Neighbors' data_knn = imp.knn(x, n_neighbors, np.mean, missing_data_cond, cat_cols) def compute_histogram(data, labels): histogram = itemfreq(sorted(data)) for label in labels: if label not in histogram[:,0]: histogram = np.vstack((histogram, np.array([[label, 0]], dtype=object))) histogram = histogram[histogram[:,0].argsort()] return histogram # compute histograms labels = np.unique(x[:,1]) freq_data = {} freq_data['Raw data'] = compute_histogram(x[:,1], labels) # freq_data['Drop missing'] = compute_histogram(data_drop[:,1], labels) freq_data['Random replace'] = compute_histogram(data_replace[:,1], labels) freq_data['Summary'] = compute_histogram(data_mode[:,1], labels) freq_data['Random forests'] = compute_histogram(data_rf[:,1], labels) freq_data['SVM'] = compute_histogram(data_svm[:,1], labels) freq_data['Logistic regression'] = compute_histogram(data_logistic[:,1], labels) freq_data['PCA'] = compute_histogram(data_facanal[:,1], labels) freq_data['KNN'] = compute_histogram(data_knn[:,1], labels) # plot histograms given feature with missing data n_methods = len(freq_data.keys()) bins = np.arange(len(labels)) width = .25 fig, ax = plt.subplots(figsize=(12,8)) for i in xrange(n_methods): key = sorted(freq_data.keys())[i] offset = i*2*width/float(n_methods) ax.bar(bins+offset, freq_data[key][:,1].astype(int), width, label=key, color=plt.cm.hot(i/float(n_methods)), align='center') ax.set_xlabel('Work class categories', size=15) ax.set_ylabel('Count', size=15) ax.set_title('Adult training set (N= 32,561)', size=15, fontweight='bold') ax.set_xticks(bins + width) ax.set_xticklabels(labels, rotation=45) plt.legend(loc=2) plt.tight_layout() plt.show()
11491834
import logging import torch import torch.nn as nn import torch.nn.functional as F from tqdm import tqdm from utils.utils import decode_mu_law from models.layers import ( UpsampleNetwork ) from models.subscale import ( CondintionNetwork, Subscaler ) from hparams import create_hparams HPARAMS = create_hparams() class WaveRNN(nn.Module): def __init__(self, hparams, debug=False): super().__init__() self.n_classes = 2 ** hparams.bits self.rnn_dims = hparams.rnn_dims self.fc_dims = hparams.fc_dims self.pad = hparams.pad self.upsample_factors = hparams.upsample_factors self.feat_dims = hparams.feat_dims self.compute_dims = hparams.compute_dims self.res_out_dims = hparams.res_out_dims self.res_blocks = hparams.res_blocks self.hop_length = hparams.hop_length self.debug = debug self.aux_dims = self.res_out_dims // 4 self.lut_x = nn.Embedding( self.n_classes, self.fc_dims, max_norm=1.0 ) self.subscale = Subscaler(hparams) self.conditioning_network = CondintionNetwork(hparams.condnet_n_layers, self.subscale.context_len, hparams.condnet_channels, hparams.condnet_kernelsize, hparams.condnet_drouput) self.upsample = UpsampleNetwork(self.feat_dims, self.upsample_factors, self.compute_dims, self.res_blocks, self.res_out_dims, self.pad) self.fc0 = nn.Linear(self.feat_dims + self.aux_dims + hparams.condnet_channels, self.rnn_dims) self.rnn1 = nn.GRU(self.rnn_dims, self.rnn_dims, batch_first=True) self.rnn2 = nn.GRU(self.rnn_dims + self.aux_dims, self.rnn_dims, batch_first=True) self.fc1 = nn.Linear(self.rnn_dims + self.aux_dims, self.fc_dims) self.fc2 = nn.Linear(self.fc_dims + self.aux_dims, self.fc_dims) self.fc3 = nn.Linear(self.fc_dims, self.n_classes) def int2float(self, x): x = 2 * x.float() / (self.n_classes - 1.) - 1. return x def forward(self, x, mel): """Method used during training. Given the chunk of a melspectrogram, and the corresponding chunk of a waveform, this decodes autoregressively, i.e. predicts softmax probabilities for the output waveform. Note: T_mel * hop_length = T_wav. args: x ([B, T_wav] torch.LongTensor): chunk of a quantized waveform (corresponds to mel) mel ([B, num_bins, T_mel] torch.FloatTensor): chunk of a mel (corresponds to x) returns: soft ([B, T_wav, 2 ** bits] torch.FloatTensor): predicted softmax probabilities. """ mel, aux = self.upsample(mel) mel, aux = self.subscale.pad(mel), self.subscale.pad(aux) aux_idx = [self.aux_dims * i for i in range(5)] a1 = aux[:, :, aux_idx[0]:aux_idx[1]] a2 = aux[:, :, aux_idx[1]:aux_idx[2]] a3 = aux[:, :, aux_idx[2]:aux_idx[3]] a4 = aux[:, :, aux_idx[3]:aux_idx[4]] x = self.int2float(x) context = self.subscale.extract_context_from_train_batch(x) # B, T, context_len -> B, context_len, T context = context.permute(0, 2, 1) conditioned = self.conditioning_network(context) # B, n_channels, T -> B, T, n_channels conditioned = conditioned.permute(0, 2, 1) x = torch.cat([conditioned, mel, a1], dim=2) x = self.subscale.stack_substensors(x) a2 = self.subscale.stack_substensors(a2) a3 = self.subscale.stack_substensors(a3) a4 = self.subscale.stack_substensors(a4) x = self.fc0(x) res = x self.rnn1.flatten_parameters() x, _ = self.rnn1(x) x = x + res res = x x = torch.cat([x, a2], dim=2) self.rnn2.flatten_parameters() x, _ = self.rnn2(x) x = x + res x = torch.cat([x, a3], dim=2) x = F.relu(self.fc1(x)) x = torch.cat([x, a4], dim=2) x = F.relu(self.fc2(x)) x = self.fc3(x) x = self.subscale.flatten_subtensors(x) soft = F.log_softmax(x, dim=-1) if self.debug: return soft, context, x return soft def train_mode_generate(self, x, mel): if self.debug: logprobs, context, x = self.forward(x, mel) else: logprobs = self.forward(x, mel) probs = torch.exp(logprobs) output = self.sampler(probs) output = self.transform(output) if self.debug: return output, context, x return output def make_context_batch(self, x, pos_dict): assert(list(pos_dict.keys()) == list(range(min(pos_dict), max(pos_dict) + 1))) context_batch = [] for subt in range(min(pos_dict), max(pos_dict) + 1): pos = pos_dict[subt] context = self.subscale.extract_context(x, pos) context_batch.append(context) context_batch = torch.stack(context_batch) context_batch = context_batch.permute(0, 2, 1) return context_batch # pylint: disable=R0913 def decode(self, context, m_t, a1_t, a2_t, a3_t, a4_t, h1, h2, rnn_cell1, rnn_cell2): """Helper function for inference mode when the ground truth waveform is not known. """ conditioned = self.conditioning_network(context) # B, n_channels, 1 -> B, n_channels conditioned = conditioned.squeeze(2) x = torch.cat([conditioned, m_t, a1_t], dim=1) x = self.fc0(x) h1 = rnn_cell1(x, h1) x = x + h1 inp = torch.cat([x, a2_t], dim=1) h2 = rnn_cell2(inp, h2) x = x + h2 x = torch.cat([x, a3_t], dim=1) x = F.relu(self.fc1(x)) x = torch.cat([x, a4_t], dim=1) x = F.relu(self.fc2(x)) x = self.fc3(x) posterior = F.softmax(1.0 * x, dim=1) if self.debug: return (posterior, h1, h2), context, x return posterior, h1, h2 # pylint: disable=R0913 def transform(self, output): output = decode_mu_law(output.float().cpu(), self.n_classes) output = output.cpu().numpy() return output def sampler(self, posterior): distrib = torch.distributions.Categorical(posterior) x = distrib.sample() x = self.int2float(x) return x def inference(self, mel, use_tqdm=True, gt=None): """Given a melspectrogram of arbitrary length, this function generates the corresponding predicted waveform. Note that T_wav = T_mel * hop_length args: mel ([B, num_bins, T_mel] torch.FloatTensor): melspectrogram. use_tqdm (bool): flag to use tqdm or not. Useful to reduce logging. returns: outputs ([B, T_wav] torch.FloatTensor): predicted waveform """ rnn_cell1 = self.get_gru_cell(self.rnn1) rnn_cell2 = self.get_gru_cell(self.rnn2) with torch.no_grad(): aux_idx = [self.aux_dims * i for i in range(5)] upsampled_mel, aux = self.upsample(mel[:, :, :]) a1 = aux[:, :, aux_idx[0]:aux_idx[1]] a2 = aux[:, :, aux_idx[1]:aux_idx[2]] a3 = aux[:, :, aux_idx[2]:aux_idx[3]] a4 = aux[:, :, aux_idx[3]:aux_idx[4]] output = torch.zeros(upsampled_mel.shape[0], upsampled_mel.shape[1]) if gt is not None: gt = gt[:, :output.shape[1]] gt = self.int2float(gt) tester_src = torch.zeros_like(output).long() tester_tgt = torch.arange(upsampled_mel.shape[1]).repeat(upsampled_mel.shape[0], 1) pos_cont = {} xs = [0] * output.shape[1] for subt in tqdm(range(self.subscale.batch_factor)): h1 = mel.new(mel.shape[0], self.rnn_dims).zero_() h2 = mel.new(mel.shape[0], self.rnn_dims).zero_() for j in tqdm(range(upsampled_mel.shape[1] // self.subscale.batch_factor)): pos = self.subscale.inv_map_pos(subt, j) m_t = upsampled_mel[:, pos, :] a1_t = a1[:, pos, :] a2_t = a2[:, pos, :] a3_t = a3[:, pos, :] a4_t = a4[:, pos, :] context = gt if gt is not None else output context = self.subscale.extract_context(context, pos) context = context.unsqueeze(1) # B, 1, context_len -> B, context_len, 1 context = context.permute(0, 2, 1) if self.debug: (posterior, h1, h2), context, x = self.decode(context, m_t, a1_t, a2_t, a3_t, a4_t, h1, h2, rnn_cell1, rnn_cell2) pos_cont[pos] = context xs[pos] = x else: posterior, h1, h2 = self.decode(context, m_t, a1_t, a2_t, a3_t, a4_t, h1, h2, rnn_cell1, rnn_cell2) sample = self.sampler(posterior) output[:, pos] = sample tester_src[:, pos] = pos assert(torch.eq(tester_src, tester_tgt).all()) output = self.transform(output) if self.debug: return output, pos_cont, torch.stack(xs, 1) return output def subscale_inference(self, mel, use_tqdm=True, gt=None): """Given a melspectrogram of arbitrary length, this function generates the corresponding predicted waveform. Note that T_wav = T_mel * hop_length args: mel ([B, num_bins, T_mel] torch.FloatTensor): melspectrogram. use_tqdm (bool): flag to use tqdm or not. Useful to reduce logging. returns: outputs ([B, T_wav] torch.FloatTensor): predicted waveform """ rnn_cell1 = self.get_gru_cell(self.rnn1) rnn_cell2 = self.get_gru_cell(self.rnn2) with torch.no_grad(): aux_idx = [self.aux_dims * i for i in range(5)] upsampled_mel, aux = self.upsample(mel[:, :, :]) a1 = aux[:, :, aux_idx[0]:aux_idx[1]] a2 = aux[:, :, aux_idx[1]:aux_idx[2]] a3 = aux[:, :, aux_idx[2]:aux_idx[3]] a4 = aux[:, :, aux_idx[3]:aux_idx[4]] output = torch.zeros(upsampled_mel.shape[0], upsampled_mel.shape[1]) if gt is not None: gt = gt[:, :output.shape[1]] gt = self.int2float(gt) tester_src = torch.zeros_like(output).long() tester_tgt = torch.arange(upsampled_mel.shape[1]).repeat(upsampled_mel.shape[0], 1) pos_cont = {} xs = [0] * output.shape[1] hidden_states = { subt: [mel.new(mel.shape[0], self.rnn_dims).zero_(), mel.new(mel.shape[0], self.rnn_dims).zero_()] for subt in range(self.subscale.batch_factor) } n_steps_subt_0 = upsampled_mel.shape[1] // self.subscale.batch_factor n_steps_remaining = (self.subscale.horizon + 1) * (self.subscale.batch_factor - 1) n_steps = n_steps_subt_0 + n_steps_remaining for j in tqdm(range(n_steps)): m_t, a1_t, a2_t, a3_t, a4_t = [], [], [], [], [] batch_size_upper_lim = min(j // (self.subscale.horizon + 1) + 1, self.subscale.batch_factor) batch_size_lower_lim = max(((j - n_steps_subt_0) // (self.subscale.horizon + 1) + 1), 0) # batch_size = batch_size_upper_lim - batch_size_lower_lim pos_dict = {} for subt in range(batch_size_lower_lim, batch_size_upper_lim): shifted_j = j - (self.subscale.horizon + 1) * subt pos = self.subscale.inv_map_pos(subt, shifted_j) pos_dict[subt] = pos m_t.append(upsampled_mel[:, pos, :]) a1_t.append(a1[:, pos, :]) a2_t.append(a2[:, pos, :]) a3_t.append(a3[:, pos, :]) a4_t.append(a4[:, pos, :]) m_t, a1_t, a2_t, a3_t, a4_t = tuple(map(lambda x: torch.stack(x).squeeze(1), [m_t, a1_t, a2_t, a3_t, a4_t])) h1 = torch.stack([hidden_states[subt][0] for subt in range(batch_size_lower_lim, batch_size_upper_lim)]).squeeze(1) h2 = torch.stack([hidden_states[subt][1] for subt in range(batch_size_lower_lim, batch_size_upper_lim)]).squeeze(1) context_source = gt if gt is not None else output context = self.make_context_batch(context_source, pos_dict) if self.debug: (posterior, h1, h2), context, x = self.decode(context, m_t, a1_t, a2_t, a3_t, a4_t, h1, h2, rnn_cell1, rnn_cell2) for i, (subt, pos) in enumerate(pos_dict.items()): pos_cont[pos] = context[i].unsqueeze(0) xs[pos] = x[i].unsqueeze(0) else: posterior, h1, h2 = self.decode(context, m_t, a1_t, a2_t, a3_t, a4_t, h1, h2, rnn_cell1, rnn_cell2) for i, subt in enumerate(range(batch_size_lower_lim, batch_size_upper_lim)): hidden_states[subt][0] = h1[i].unsqueeze(0) hidden_states[subt][1] = h2[i].unsqueeze(0) sample = self.sampler(posterior) for i, (subt, pos) in enumerate(pos_dict.items()): output[:, pos] = sample[i] tester_src[:, pos] = pos assert(torch.eq(tester_src, tester_tgt).all()) output = self.transform(output) if self.debug: return output, pos_cont, torch.stack(xs, 1) return output @staticmethod def get_gru_cell(gru): """Given a GRU object, this function returns the corresponding GRU cell with the correct weights initialised from the GRU object. args: gru (nn.GRU): GRU from which to get the cell. returns: gru_cell (nn.GRUCell): GRU cell where the cell has been initialised from gru. """ gru_cell = nn.GRUCell(gru.input_size, gru.hidden_size) gru_cell.weight_hh.data = gru.weight_hh_l0.data gru_cell.weight_ih.data = gru.weight_ih_l0.data gru_cell.bias_hh.data = gru.bias_hh_l0.data gru_cell.bias_ih.data = gru.bias_ih_l0.data return gru_cell def load_max_state_dict(self, state_dict): own_state = self.state_dict() for name, param in state_dict.items(): if name not in own_state: logging.warning("This weight is not in model: {}".format(name)) continue try: own_state[name].copy_(param) except RuntimeError: logging.warning(name) logging.warning("Model size: {}".format(own_state[name].size())) logging.warning("Checkpoint size: {}".format(param.size()))
11491841
import torch as th from typing import Optional, Tuple from tpp.utils import batch as bu from tpp.utils.events import Events from tpp.utils.utils import smallest_positive from tpp.utils.index import take_2_by_2 from tpp.utils.history_bst import get_prev_times as get_prev_times_bst def _get_rank(x: th.Tensor) -> int: return len(x.shape) def expand_to_rank(x: th.Tensor, rank: int, dim: int = -1) -> th.Tensor: """Expand a tensor to a desired rank. Args: x: The tensor to expand. rank: The target rank. dim: The dim to expand along. Defaults to `-1`. Returns: A tensor expanded to the given rank. """ x_rank = _get_rank(x) if x_rank > rank: raise ValueError( "Rank of `x` ({}) greater than desired rank ({})".format( x_rank, rank)) for _ in range(rank - x_rank): x = x.unsqueeze(dim) return x def build_histories( query: th.Tensor, events: Events, history_size: Optional[int] = 1) -> Tuple[th.Tensor, th.Tensor]: """Get the set of times corresponding to the 'history' of a query time of fixed size. Args: query: [B,T] The times to create histories for. events: [B,L] Times and labels of events to create histories from. history_size: The size of each history. Defaults to 1. Returns: history (th.Tensor): [B,T,H] The history for each query time. mask (th.Tensor): [B,T] The mask corresponding to whether a particular query can be used or not based on the required size of history. """ batch_size, max_queries = query.shape batch_size_s, max_seq_len = events.times.shape if batch_size_s != batch_size: raise ValueError( "The batch size for `query_times` " "({}) does not match the batch size for `sequences` " "({}).".format(batch_size, batch_size_s)) if history_size > max_seq_len: raise ValueError( "The chosen value for `history_size` " "({}) is greater than the size of the largest sequence " "({}).".format(history_size, max_seq_len)) ((prev_times, prev_times_idxs), is_event, pos_delta_mask) = get_prev_times_bst( query=query, events=events) # ([B,T], [B,T]), [B,T], [B,T] relative_history_idxs = th.arange( start=1 - history_size, end=1, device=events.times.device) batch_idxs_shift = th.arange( start=0, end=batch_size_s, device=events.times.device) * max_seq_len batch_idxs_shift = batch_idxs_shift.reshape([batch_size, 1, 1]) history_seq_idxs = prev_times_idxs.reshape([batch_size, max_queries, 1]) history_seq_idxs = history_seq_idxs + relative_history_idxs batch_history_seq_idxs = history_seq_idxs + batch_idxs_shift batch_history_seq_idxs = batch_history_seq_idxs.long() # [B,T,H] history = th.take(events.times, batch_history_seq_idxs) history_idxs_positive = th.prod(history_seq_idxs >= 0, dim=-1) # [B,T] history_idxs_positive = history_idxs_positive.type(pos_delta_mask.dtype) history_mask = pos_delta_mask * history_idxs_positive history_mask = history_mask.type(history.dtype) # [B,T] return history, history_mask def get_prev_times( query: th.Tensor, events: Events, allow_window: Optional[bool] = False ) -> Tuple[Tuple[th.Tensor, th.Tensor], th.Tensor, th.Tensor]: """For each query, get the event time that directly precedes it. If no events precedes it (but the window start does), return the window start. Otherwise, mask the value. Args: query: [B,T] Sequences of query times to evaluate the intensity function. events: [B,L] Times and labels of events. allow_window: If `True`, a previous time can be the window boundary. Defaults to `False`. Returns: `times` is a tuple of tensor of values [B,T] and indices, [B,T] of the largest time value in the sequence that is strictly smaller than the query time value, or the window. the index only event indexes into the events. If the window is returned, it should be dealt with explicitly at encoding/decoding time. `is_event` is a tensor [B,T] that indicates whether the time corresponds to an event or not (a 1 indicates an event and a 0 indicates a window boundary). `mask` is a tensor [B,T] that indicates whether the time difference to those times what positive or not. """ event_times = events.get_times(prepend_window=allow_window) # [B,L+1] batch_size, max_seq_len = event_times.shape time_diffs = bu.batchwise_difference(query, event_times) # [B,T,L+1] event_mask = events.get_mask(prepend_window=allow_window) # [B,L+1] event_mask = event_mask.reshape([batch_size, 1, max_seq_len]) # [B,1,L+1] time_diffs = time_diffs * event_mask # [B,T,L+1] smallest_time_diffs, mask = smallest_positive(time_diffs, dim=-1) # [B,T] prev_times_idxs = smallest_time_diffs.indices # [B,T] prev_times = take_2_by_2(event_times, index=prev_times_idxs) # [B,T] if allow_window: # If the first event shares a time with the window boundary, that the # index returned is the index of the event, rather than the window # boundary. idx_is_window = (prev_times_idxs == 0).type( prev_times_idxs.dtype) # [B,T] do_idx_shift = events.first_event_on_window.type( idx_is_window.dtype) # [B] idx_shift = idx_is_window * do_idx_shift.reshape(-1, 1) prev_times_idxs = prev_times_idxs + idx_shift # Check the indexes in case one of the window indexes became an event. is_event = (prev_times_idxs != 0).type(mask.dtype) # [B,T] else: is_event = th.ones_like(prev_times_idxs) # [B,T] query_above_window = query > events.window_start.reshape(-1, 1) query_below_window = query <= events.window_end.reshape(-1, 1) query_within_window = query_above_window & query_below_window query_within_window = query_within_window.type(mask.dtype) mask = mask * query_within_window return (prev_times, prev_times_idxs), is_event, mask # [B,T]
11491859
import os import pathlib from pytest_notebook.nb_regression import NBRegressionFixture EXEC_CWD = str(pathlib.Path(__file__).resolve().parent.parent) fixture = NBRegressionFixture( exec_timeout=120, exec_cwd=EXEC_CWD, diff_color_words=True, diff_ignore=( "/cells/*/outputs/*/data/text/plain", "/cells/*/outputs/*/data/image/svg+xml", "/cells/*/outputs/*/text", "/cells/*/outputs/", "/cells/*/outputs/*/data/image/png", "/cells/*/outputs/*/data/text/html", "/cells/*/outputs/*/output/data/" "/cells/*/outputs/*/data/application/vnd.plotly.v1+json", "/cells/*/outputs/*/execution_count", "/cells/*/execution_count", "/cells/1/outputs/", "/cells/*/execution_count/", "/cells/*/metadata/", "/cells/*/outputs/metadata/", "/metadata/", ), ) def test_cluster_analysis_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Cluster_Analysis.ipynb") fixture.check(notebook, raise_errors=True) def test_data_summary_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Data_Summary.ipynb") fixture.check(notebook, raise_errors=True) def test_text_preprocessing_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Text_Preprocessing.ipynb") fixture.check(notebook, raise_errors=True) def test_data_heatmap_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Data_Heatmap.ipynb") fixture.check(notebook, raise_errors=True) def test_feature_importance_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Feature_Importance.ipynb") fixture.check(notebook, raise_errors=True) def test_correlation_matrix_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Correlation_Matrix.ipynb") fixture.check(notebook, raise_errors=True) def test_scatter_plots_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Scatter_Plots.ipynb") fixture.check(notebook, raise_errors=True) def test_tutorial_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Quick_Start.ipynb") fixture.check(notebook, raise_errors=True) def test_topic_modeling_notebook(): notebook = os.path.join(EXEC_CWD, "examples", "Topic_Modeling.ipynb") fixture.check(notebook, raise_errors=True)
11491983
from toee import * from utilities import * from py00439script_daemon import * from combat_standard_routines import * def san_start_combat( attachee, triggerer ): if (attachee.leader_get() != OBJ_HANDLE_NULL and not npc_get(attachee,2)): leader = attachee.leader_get() if (group_pc_percent_hp( attachee, leader ) <= 40): attachee.obj_set_int(obj_f_critter_strategy, 462) elif (game.party_npc_size() + game.party_pc_size() == 8): for pp in range(0,8): if (game.party[pp] != OBJ_HANDLE_NULL): if (obj_percent_hp(game.party[pp]) <= 50 and game.party[pp].stat_level_get(stat_hp_current) >= -9): game.global_flags[250 + pp] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[252] == 1): if (adjacent(attachee, game.party[2])): game.global_flags[259] = 1 if (game.global_flags[253] == 1): if (adjacent(attachee, game.party[3])): game.global_flags[259] = 1 if (game.global_flags[254] == 1): if (adjacent(attachee, game.party[4])): game.global_flags[259] = 1 if (game.global_flags[255] == 1): if (adjacent(attachee, game.party[5])): game.global_flags[259] = 1 if (game.global_flags[256] == 1): if (adjacent(attachee, game.party[6])): game.global_flags[259] = 1 if (game.global_flags[257] == 1): if (adjacent(attachee, game.party[7])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_npc_size() + game.party_pc_size() == 7): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[1]) <= 50 and game.party[1].stat_level_get(stat_hp_current) >= -9): game.global_flags[251] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[2]) <= 50 and game.party[2].stat_level_get(stat_hp_current) >= -9): game.global_flags[252] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[3]) <= 50 and game.party[3].stat_level_get(stat_hp_current) >= -9): game.global_flags[253] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[4]) <= 50 and game.party[4].stat_level_get(stat_hp_current) >= -9): game.global_flags[254] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[5]) <= 50 and game.party[5].stat_level_get(stat_hp_current) >= -9): game.global_flags[255] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[6]) <= 50 and game.party[6].stat_level_get(stat_hp_current) >= -9): game.global_flags[256] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[252] == 1): if (adjacent(attachee, game.party[2])): game.global_flags[259] = 1 if (game.global_flags[253] == 1): if (adjacent(attachee, game.party[3])): game.global_flags[259] = 1 if (game.global_flags[254] == 1): if (adjacent(attachee, game.party[4])): game.global_flags[259] = 1 if (game.global_flags[255] == 1): if (adjacent(attachee, game.party[5])): game.global_flags[259] = 1 if (game.global_flags[256] == 1): if (adjacent(attachee, game.party[6])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_npc_size() + game.party_pc_size() == 6): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[1]) <= 50 and game.party[1].stat_level_get(stat_hp_current) >= -9): game.global_flags[251] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[2]) <= 50 and game.party[2].stat_level_get(stat_hp_current) >= -9): game.global_flags[252] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[3]) <= 50 and game.party[3].stat_level_get(stat_hp_current) >= -9): game.global_flags[253] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[4]) <= 50 and game.party[4].stat_level_get(stat_hp_current) >= -9): game.global_flags[254] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[5]) <= 50 and game.party[5].stat_level_get(stat_hp_current) >= -9): game.global_flags[255] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[252] == 1): if (adjacent(attachee, game.party[2])): game.global_flags[259] = 1 if (game.global_flags[253] == 1): if (adjacent(attachee, game.party[3])): game.global_flags[259] = 1 if (game.global_flags[254] == 1): if (adjacent(attachee, game.party[4])): game.global_flags[259] = 1 if (game.global_flags[255] == 1): if (adjacent(attachee, game.party[5])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_npc_size() + game.party_pc_size() == 5): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[1]) <= 50 and game.party[1].stat_level_get(stat_hp_current) >= -9): game.global_flags[251] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[2]) <= 50 and game.party[2].stat_level_get(stat_hp_current) >= -9): game.global_flags[252] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[3]) <= 50 and game.party[3].stat_level_get(stat_hp_current) >= -9): game.global_flags[253] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[4]) <= 50 and game.party[4].stat_level_get(stat_hp_current) >= -9): game.global_flags[254] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[252] == 1): if (adjacent(attachee, game.party[2])): game.global_flags[259] = 1 if (game.global_flags[253] == 1): if (adjacent(attachee, game.party[3])): game.global_flags[259] = 1 if (game.global_flags[254] == 1): if (adjacent(attachee, game.party[4])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_npc_size() + game.party_pc_size() == 4): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[1]) <= 50 and game.party[1].stat_level_get(stat_hp_current) >= -9): game.global_flags[251] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[2]) <= 50 and game.party[2].stat_level_get(stat_hp_current) >= -9): game.global_flags[252] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[3]) <= 50 and game.party[3].stat_level_get(stat_hp_current) >= -9): game.global_flags[253] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[252] == 1): if (adjacent(attachee, game.party[2])): game.global_flags[259] = 1 if (game.global_flags[253] == 1): if (adjacent(attachee, game.party[3])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_npc_size() + game.party_pc_size() == 3): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[1]) <= 50 and game.party[1].stat_level_get(stat_hp_current) >= -9): game.global_flags[251] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[2]) <= 50 and game.party[2].stat_level_get(stat_hp_current) >= -9): game.global_flags[252] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[252] == 1): if (adjacent(attachee, game.party[2])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_npc_size() + game.party_pc_size() == 2): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (obj_percent_hp(game.party[1]) <= 50 and game.party[1].stat_level_get(stat_hp_current) >= -9): game.global_flags[251] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[251] == 1): if (adjacent(attachee, game.party[1])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) elif (game.party_pc_size() == 1): if (obj_percent_hp(game.party[0]) <= 50 and game.party[0].stat_level_get(stat_hp_current) >= -9): game.global_flags[250] = 1 game.global_flags[258] = 1 if (game.global_flags[250] == 1): if (adjacent(attachee, game.party[0])): game.global_flags[259] = 1 if (game.global_flags[258] == 1): if (game.global_flags[259] == 1): attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) else: attachee.obj_set_int(obj_f_critter_strategy, 463) else: attachee.obj_set_int(obj_f_critter_strategy, 464) if (triggerer.type == obj_t_npc and triggerer.leader_get() == OBJ_HANDLE_NULL): attachee.turn_towards(triggerer) else: for pc in game.party: if ( pc.has_feat(feat_animal_companion) ): attachee.turn_towards(pc) else: attachee.turn_towards(game.party[0]) game.global_flags[250] = 0 game.global_flags[251] = 0 game.global_flags[252] = 0 game.global_flags[253] = 0 game.global_flags[254] = 0 game.global_flags[255] = 0 game.global_flags[256] = 0 game.global_flags[257] = 0 game.global_flags[258] = 0 game.global_flags[259] = 0 return RUN_DEFAULT def san_join( attachee, triggerer ): if (npc_get(attachee,1)): npc_set(attachee,2) return RUN_DEFAULT def san_spell_cast( attachee, triggerer, spell ): if ( spell.spell == spell_charm_person_or_animal or spell.spell == spell_charm_monster ): npc_set(attachee,1) return RUN_DEFAULT def not_adjacent( companion, target ): if (companion.distance_to(target) >= 5): return 1 return 0 def adjacent( companion, target ): if (companion.distance_to(target) <= 5): return 1 return 0
11491984
import os import textwrap class CodeGenerator(): def __init__(self): self.layers = [] def addLayer(self, weights, biases, activation): self.layers.append((weights, biases, activation)) def save(self, name): lines = [] lines.append('// Auto generated model data') lines.append('private void InitModel()') lines.append('{') numInputs = self.layers[0][0].shape[1] lines.append('\tNumInputs = %d;' % numInputs) numOutputs = [str(layer[0].shape[0]) for layer in self.layers] lines.append('\tNumOutputs = {%s};' % ', '.join(numOutputs)) for i in range(len(self.layers)): lines.append('\tLayer%d_W = %s;' % (i, self.convertToCSArray(self.layers[i][0]))) lines.append('\tLayer%d_B = %s;' % (i, self.convertToCSArray(self.layers[i][1]))) lines.append('}') file = open(name, 'w') for line in lines: wrapLines = textwrap.wrap(line, width=120, tabsize=4, subsequent_indent='\t\t') for fileLine in wrapLines: file.write(fileLine + '\n') file.close() print('Saved generated code:',name) def convertToCSArray(self, arr): values = [] if len(arr.shape) == 1: values = [str(x) for x in arr] else: for i in range(arr.shape[0]): values.append(self.convertToCSArray(arr[i])) return '{' + ', '.join(values) + '}'
11491987
import argparse import subprocess import time def track(wait_time=60): # In the following, a select set of parameters that are passed to ffmpeg is # described: # -vframes -> Number of frames to output. # -q:v -> (alias: qscale) Quality of the output, 100 means as lossy as # possible (resulting in less filesize). # -vf scale=1280:-1 -> Rescale to 1280 width (and automatically keep aspect # ratio). Uses ffmpeg filtering. args = ("ffmpeg -f avfoundation -framerate 1 -i 1:0 -vframes 1 -q:v 100 " "-vf scale=1280:-1") args = args.split(" ") while True: time.sleep(wait_time) filename = "output" + str(int(time.time())) + ".jpeg" args_ = args.copy() args_.append(filename) p = subprocess.Popen(args_) p.wait() def watch(): pass # TODO if __name__ == '__main__': parser = argparse.ArgumentParser() parser.add_argument( 'cmd', nargs='?', default='track', help='Command to execute. Can be either track or watch.' ) parser.add_argument( '--waittime', help='How many seconds to wait between each capture.', default=60, type=int ) args = parser.parse_args() if args.cmd == 'track': track(args.waittime) elif args.cmd == 'watch': watch()
11491990
import collections from setuptools import Extension from buildtools import * import torch import os # ~~~ libnitorch files # Note that sources are identical between libnitorch_cpu and libnitorch_cuda. # The same code is compiled in two different ways to generate native and cuda # code. This trick allows to minimize code duplication. # Finally, libnitorch links to both these sub-libraries and dispatches # according to the Tensor's device type. libnitorch_cpu_sources = ['pushpull_common.cpp'] libnitorch_cuda_sources = ['pushpull_common.cpp'] libnitorch_sources = ['pushpull.cpp'] ext_spatial_sources = ['spatial.cpp'] # That's a bit ugly. TODO: use config files? libnitorch_cpu_headers = ['common.h', 'interpolation.h', 'interpolation_common.h', 'bounds.h', 'bounds_common.h'] libnitorch_cuda_headers = ['common.h', 'interpolation.h', 'interpolation_common.h', 'bounds.h', 'bounds_common.h'] libnitorch_headers = ['pushpull_common.h', 'interpolation.h', 'bounds.h'] ext_spatial_headers = ['pushpull.h', 'interpolation.h', 'bounds.h'] # TODO # . There is still quite a lot to do in setup and buildtools in order to make # things clean and work on multiple platforms. # . I have to add abi checks and other smart tricks as in # torch.utils.cpp_extension MINIMUM_GCC_VERSION = (5, 0, 0) MINIMUM_MSVC_VERSION = (19, 0, 24215) # ~~~ helpers # Most of the helpers are in build tools. The remaining helpers defined # here are specific to the version of pytorch that we compile against. def torch_version(astuple=True): version = list(torch.__version__.split('+')[0].split('.')) # strip alpha tags for n, v in enumerate(version): for x in 'abcdefghijklmnopqrstuvwxy': if x in v: v = v[:v.index(x)] version[n] = v version = tuple(int(v) for v in version) if len(version) == 2: version = version + (0,) if not astuple: version = version[0]*10000 + version[1]*100 + version[0] return version def torch_cuda_version(astuple=True): version = torch.version.cuda.split('.') version = tuple(int(v) for v in version) if len(version) == 2: version = version + (0,) if not astuple: version = version[0]*10000 + version[1]*100 + version[0] return version def torch_cudnn_version(astuple=True): version = torch.backends.cudnn.version() version = (version//1000, version//100 % 10, version % 100) if not astuple: version = version[0]*10000 + version[1]*100 + version[0] return version def torch_parallel_backend(): # check if set by user valid_backends = ('AT_PARALLEL_OPENMP', 'AT_PARALLEL_NATIVE', 'AT_PARALLEL_NATIVE_TBB') backend = os.environ.get('NI_PARALLEL_BACKEND', None) if backend: if backend not in valid_backends: backend = None return backend # else, find backend used by pytorch match = re.search('^ATen parallel backend: (?P<backend>.*)$', torch._C._parallel_info(), re.MULTILINE) if match is None: return None backend = match.group('backend') if backend == 'OpenMP': return 'AT_PARALLEL_OPENMP' elif backend == 'native thread pool': return 'AT_PARALLEL_NATIVE' elif backend == 'native thread pool and TBB': return 'AT_PARALLEL_NATIVE_TBB' else: return None def torch_abi(): return str(int(torch._C._GLIBCXX_USE_CXX11_ABI)) def torch_omp_lib(): torch_dir = os.path.dirname(os.path.abspath(torch.__file__)) torch_library_dir = os.path.join(torch_dir, 'lib') if is_darwin(): libtorch = os.path.join(torch_library_dir, 'libtorch.dylib') linked_libs = os.popen('otool -L "{}"'.format(libtorch)).read() if 'libiomp5' in linked_libs: return 'iomp5' elif 'libomp' in linked_libs: return 'omp' else: return None def torch_libraries(use_cuda=False): version = torch_version(astuple=False) if version < 10500: libraries = ['c10', 'torch', 'torch_python'] if use_cuda: libraries += ['cudart', 'c10_cuda'] else: libraries = ['c10', 'torch_cpu', 'torch_python', 'torch'] if use_cuda: libraries += ['cudart', 'c10_cuda', 'torch_cuda'] if not use_cuda and torch_parallel_backend() == 'AT_PARALLEL_OPENMP': libraries += omp_libraries() return libraries def torch_library_dirs(use_cuda=False, use_cudnn=False): torch_dir = os.path.dirname(os.path.abspath(torch.__file__)) torch_library_dir = os.path.join(torch_dir, 'lib') library_dirs = [torch_library_dir] if use_cuda: if is_windows(): library_dirs += [os.path.join(cuda_home(), 'lib/x64')] elif os.path.exists(os.path.join(cuda_home(), 'lib64')): library_dirs += [os.path.join(cuda_home(), 'lib64')] elif os.path.exists(os.path.join(cuda_home(), 'lib')): library_dirs += [os.path.join(cuda_home(), 'lib')] if use_cudnn: if is_windows(): library_dirs += [os.path.join(cudnn_home(), 'lib/x64')] elif os.path.exists(os.path.join(cudnn_home(), 'lib64')): library_dirs += [os.path.join(cudnn_home(), 'lib64')] elif os.path.exists(os.path.join(cudnn_home(), 'lib')): library_dirs += [os.path.join(cudnn_home(), 'lib')] if not use_cuda and torch_parallel_backend() == 'AT_PARALLEL_OPENMP': library_dirs += omp_library_dirs() return library_dirs def torch_include_dirs(use_cuda=False, use_cudnn=False): torch_dir = os.path.dirname(os.path.abspath(torch.__file__)) torch_include_dir = os.path.join(torch_dir, 'include') include_dirs = [torch_include_dir, os.path.join(torch_include_dir, 'torch', 'csrc', 'api', 'include'), os.path.join(torch_include_dir, 'TH'), os.path.join(torch_include_dir, 'THC')] if use_cuda: cuda_include_dir = os.path.join(cuda_home(), 'include') if cuda_include_dir != '/usr/include': include_dirs += [cuda_include_dir] if use_cudnn: include_dirs += [os.path.join(cudnn_home(), 'include')] if not use_cuda and torch_parallel_backend() == 'AT_PARALLEL_OPENMP': include_dirs += omp_include_dirs() return include_dirs def cuda_check(): local_version = cuda_version() torch_version = torch_cuda_version() ok = (local_version[0] == torch_version[0] and local_version[1] == torch_version[1]) if not ok: print('Your version of CUDA is v{}.{} while PyTorch was compiled with' 'CUDA v{}.{}. NiTorch cannot be compiled with CUDA.'.format( local_version[0], local_version[1], torch_version[0], torch_version[1])) return ok def cudnn_check(): local_version = cudnn_version() torch_version = torch_cudnn_version() ok = (local_version[0] == torch_version[0] and local_version[1] == torch_version[1]) if not ok: print('Your version of CuDNN is v{}.{} while PyTorch was compiled with' 'CuDNN v{}.{}. NiTorch cannot be compiled with CuDNN.'.format( local_version[0], local_version[1], torch_version[0], torch_version[1])) return ok def cuda_arch_flags(): """ Determine CUDA arch flags to use. For an arch, say "6.1", the added compile flag will be ``-gencode=arch=compute_61,code=sm_61``. For an added "+PTX", an additional ``-gencode=arch=compute_xx,code=compute_xx`` is added. See select_compute_arch.cmake for corresponding named and supported arches when building with CMake. """ # Note: keep combined names ("arch1+arch2") above single names, otherwise # string replacement may not do the right thing named_arches = collections.OrderedDict([ ('Kepler+Tesla', '3.7'), ('Kepler', '3.5+PTX'), ('Maxwell+Tegra', '5.3'), ('Maxwell', '5.0;5.2+PTX'), ('Pascal', '6.0;6.1+PTX'), ('Volta', '7.0+PTX'), ('Turing', '7.5+PTX'), ('Ampere', '8.0;8.6+PTX'), ]) supported_arches = ['3.5', '3.7', '5.0', '5.2', '5.3', '6.0', '6.1', '6.2', '7.0', '7.2', '7.5', '8.0', '8.6'] valid_arch_strings = supported_arches + [s + "+PTX" for s in supported_arches] # The default is sm_30 for CUDA 9.x and 10.x # First check for an env var (same as used by the main setup.py) # Can be one or more architectures, e.g. "6.1" or "3.5;5.2;6.0;6.1;7.0+PTX" # See cmake/Modules_CUDA_fix/upstream/FindCUDA/select_compute_arch.cmake arch_list = os.environ.get('TORCH_CUDA_ARCH_LIST', 'mine') # If not given, look into libtorch_cuda if not arch_list or arch_list.lower() == 'all': cuobjdump = os.path.join(cuda_home(), 'bin', 'cuobjdump') torchdir = os.path.dirname(os.path.abspath(torch.__file__)) libtorch = os.path.join(torchdir, 'lib') if is_windows(): libtorch = os.path.join(libtorch, 'torch_cuda.lib') else: assert not is_darwin() libtorch = os.path.join(libtorch, 'libtorch_cuda.so') arch_list = os.popen(cuobjdump + " '" + libtorch + \ "' -lelf | awk -F. '{print $3}' | " \ "grep sm | sort -u").read().split('\n') arch_list = [arch[3] + '.' + arch[4] for arch in arch_list if arch] ptx_list = os.popen(cuobjdump + " '" + libtorch + \ "' -lptx | awk -F. '{print $3}' | " \ "grep sm | sort -u").read().split('\n') ptx_list = [arch[3] + '.' + arch[4] for arch in ptx_list if arch] arch_list = [arch + '+PTX' if arch in ptx_list else arch for arch in arch_list] elif arch_list == 'mine': # this bit was in the torch extension util but I have replaced # it with the bit above that looks into libtorch capability = torch.cuda.get_device_capability() arch_list = ['{}.{}'.format(capability[0], capability[1])] else: # Deal with lists that are ' ' separated (only deal with ';' after) arch_list = arch_list.replace(' ', ';') # Expand named arches for named_arch, archval in named_arches.items(): arch_list = arch_list.replace(named_arch, archval) arch_list = arch_list.split(';') flags = [] for arch in arch_list: if arch not in valid_arch_strings: raise ValueError("Unknown CUDA arch ({}) or GPU not supported".format(arch)) else: num = arch[0] + arch[2] flags.append('-gencode=arch=compute_{},code=sm_{}'.format(num, num)) if arch.endswith('+PTX'): flags.append('-gencode=arch=compute_{},code=compute_{}'.format(num, num)) return list(set(flags)) def torch_extension_flags(name): return ['-DTORCH_EXTENSION_NAME={}'.format(name), '-DTORCH_API_INCLUDE_EXTENSION_H'] def gcc_clang_flags(): flags = ['-fPIC', '-std=c++14'] if is_darwin() and darwin_cc_type() == 'apple_clang': flags += ['-stdlib=libc++'] return flags def msvc_flags(): return ['/MD', '/wd4819', '/EHsc'] def nvcc_flags(): return [ '-x=cu', '-D__CUDA_NO_HALF_OPERATORS__', '-D__CUDA_NO_HALF_CONVERSIONS__', '-D__CUDA_NO_HALF2_OPERATORS__', '--expt-relaxed-constexpr'] def darwin_cc_type(): CC = os.environ.get('CC', 'clang') CC_name = os.popen(CC + ' --version').read().split(' ')[0] if CC_name == 'Apple': CC_type = 'apple_clang' elif CC_name == 'clang': CC_type = 'other_clang' else: CC_type = 'other' return CC_type def find_omp_darwin(): """Set the correct openmp flag on MacOS. LLVM's clang has both GCC's and Intel's implementations of OpenMP (gomp, omp5), which can be specifically used with the flag '-fopenmp=libiomp5' or '-fopenmp=libgomp'. If no implementation is specified, the behaviour depends on the version of LLVM (it used to be gomp, now I believe it is iomp5). Note that iomp5 is binary compatible with gomp (does that mean we can link with both without conflicts?). Apple's clang does not ship any OpenMP implementation, so the user needs to install one herself, which we need to find and reference properly. Return (cflag, lflag, lib_name, lib_dir).""" # TODO: LLVM's clang embeds OpenMP for version >= 3.8.0 # I need to add a special case for earlier versions. # There are other ABI incompatibilities as well: # https://openmp.llvm.org # Various references that helped me in this maze: # https://iscinumpy.gitlab.io/post/omp-on-high-sierra/ # https://stackoverflow.com/questions/37362414/ # https://reviews.llvm.org/D2841 def find_lib(names): for name in names: if not name: continue dirs = ['.', '/usr/', '/usr/local/', '/opt/local/', '/usr/local/opt/libomp/'] if os.environ.get('LD_LIBRARY_PATH'): dirs += os.environ.get('LD_LIBRARY_PATH').split(':') for dir in dirs: if os.path.exists(os.path.join(dir, 'lib', 'lib' + name + '.dylib')): return name, dir return None, None # First, check which clang we're dealing with # (gcc, apple clang or external clang) CC_type = darwin_cc_type() # If not apple clang: openmp should be packaged with the compiler: if CC_type != 'apple_clang': flag = '-fopenmp' if CC_type == 'other_clang': if torch_omp_lib() == 'iomp5': flag = '-fopenmp=libiomp5' elif torch_omp_lib() == 'omp': flag = '-fopenmp=libgomp' return [flag], [flag], [], None # Else, opnemp is no different than any other dependency # First, check if omp/iomp5 has been installed (e.g., using homebrew) lib_name, lib_dir = find_lib([torch_omp_lib(), 'iomp5', 'omp']) if lib_name is None: # OpenMP not found. # Let's just hope that the compiler knows what it's doing. return ['-fopenmp'], ['-fopenmp'], [], None else: return ['-Xpreprocessor', '-fopenmp'], [], [], lib_dir # return ['-Xpreprocessor', '-fopenmp'], [lib_name], [], lib_dir # It is super weird: # - precompiled torch wheels on mac link against libomp (or libiomp5) # - but openmp was not detected in their compilation chain, so # it was actually compiled *without* openmp (pragmas were not used, # parallel loops are actually sequential, (set/get)_num_threads # is always 1) # - I can't get my mac to link against the correct omp lib (it # links against libomp even though I ask it to link against # libiomp5) # - If I don't link against openmp at all, things seem to work (!!) # - I really need to rewrite our compilation stuff anyway def omp_flags(): if is_windows(): return ['/openmp'] elif is_darwin(): return find_omp_darwin()[0] else: return ['-fopenmp'] def omp_link_flags(): if is_darwin(): return find_omp_darwin()[1] else: return [] def omp_libraries(): if is_darwin(): return find_omp_darwin()[2] else: return [] def omp_library_dirs(): if is_darwin(): ompdir = find_omp_darwin()[3] return [os.path.join(ompdir, 'lib')] if ompdir else [] else: return [] def omp_include_dirs(): if is_darwin(): ompdir = find_omp_darwin()[3] return [os.path.join(ompdir, 'include')] if ompdir else [] else: return [] def common_flags(): if is_windows(): return msvc_flags() else: return gcc_clang_flags() def torch_flags(cuda=False): version = torch_version() version = version[0]*10000+version[1]*100+version[2] flags = ['-DNI_TORCH_VERSION=' + str(version)] backend = torch_parallel_backend() flags += [ '-D' + torch_parallel_backend() + '=1', '-D_GLIBCXX_USE_CXX11_ABI=' + torch_abi()] if not cuda and backend == 'AT_PARALLEL_OPENMP': flags += omp_flags() return flags def torch_link_flags(cuda=False): backend = torch_parallel_backend() flags = [] if not cuda and backend == 'AT_PARALLEL_OPENMP': flags += omp_link_flags() return flags def common_links_flags(): if is_darwin() and darwin_cc_type() == 'apple_clang': return ['-stdlib=libc++'] return [] def cuda_flags(): flags = nvcc_flags() + cuda_arch_flags() if is_windows(): for flag in common_flags(): flags = ['-Xcompiler', flag] + flags else: for flag in common_flags(): flags += ['--compiler-options', flag] return flags def abspathC(files): scriptdir = os.path.abspath(os.path.dirname(__file__)) sourcedir = os.path.join(scriptdir, 'nitorch', '_C') return [os.path.join(sourcedir, f) for f in files] def prepare_extensions(): build_extensions = [] # ~~~ checks use_cuda = bool(int(os.environ.get('NI_USE_CUDA', '1'))) use_cuda = use_cuda and cuda_home() and cuda_check() use_cudnn = False # cudnn_home() and cudnn_check() nitorch_lib = [] nitorch_libext = [] # ~~~ setup libraries NiTorchCPULibrary = SharedLibrary( name='lib.nitorch_cpu', sources=abspathC(libnitorch_cpu_sources), depends=abspathC(libnitorch_cpu_headers), libraries=torch_libraries(), library_dirs=torch_library_dirs(), include_dirs=torch_include_dirs(), extra_compile_args=common_flags() + torch_flags(), extra_link_args=common_links_flags() + torch_link_flags(), language='c++', ) build_extensions += [NiTorchCPULibrary] nitorch_libext += [NiTorchCPULibrary] nitorch_lib += ['nitorch_cpu'] if use_cuda: NiTorchCUDALibrary = SharedLibrary( name='lib.nitorch_cuda', sources=abspathC(libnitorch_cuda_sources), depends=abspathC(libnitorch_cuda_headers), libraries=torch_libraries(use_cuda), library_dirs=torch_library_dirs(use_cuda, use_cudnn), include_dirs=torch_include_dirs(use_cuda, use_cudnn), extra_compile_args=cuda_flags() + torch_flags(cuda=True), extra_link_args=torch_link_flags(cuda=True), language='cuda', ) build_extensions += [NiTorchCUDALibrary] nitorch_libext += [NiTorchCUDALibrary] nitorch_lib += ['nitorch_cuda'] NiTorchLibrary = SharedLibrary( name='lib.nitorch', sources=abspathC(libnitorch_sources), depends=nitorch_libext + abspathC(libnitorch_headers), libraries=torch_libraries() + nitorch_lib, library_dirs=torch_library_dirs(), include_dirs=torch_include_dirs(), extra_compile_args=common_flags() + torch_flags() + (['-DNI_WITH_CUDA'] if use_cuda else []), extra_link_args=common_links_flags(), runtime_library_dirs=[link_relative('.')], language='c++', ) build_extensions += [NiTorchLibrary] nitorch_libext = [NiTorchLibrary] nitorch_lib = ['nitorch'] # ~~~ setup extensions python_library_dirs = [os.path.join(sys.exec_prefix, 'lib')] SpatialExtension = Extension( name='_C.spatial', sources=abspathC(ext_spatial_sources), depends=nitorch_libext + abspathC(ext_spatial_headers), libraries=torch_libraries(use_cuda) + nitorch_lib, library_dirs=torch_library_dirs(use_cuda, use_cudnn) + python_library_dirs, include_dirs=torch_include_dirs(use_cuda, use_cudnn), extra_compile_args=common_flags() + torch_flags() + torch_extension_flags('spatial'), extra_link_args=common_links_flags(), runtime_library_dirs=[link_relative(os.path.join('..', 'lib'))] ) build_extensions += [SpatialExtension] return build_extensions
11492066
from typing import Optional import re from collections import OrderedDict from requests import Response as ServerResponse from instagram_api.response.mapper import ApiResponse from .account_disabled import AccountDisabledException from .bad_request import BadRequestException from .challenge_required import ChallengeRequiredException from .checkpoint_required import CheckpointRequiredException from .consent_required import ConsentRequiredException from .endpoint import EndpointException from .feedback_required import FeedbackRequiredException from .forced_password_reset import ForcedPasswordResetException from .incorrect_password import IncorrectPasswordException from .instagram import InstagramException from .invalid_sms_code import InvalidSmsCodeException from .invalid_user import InvalidUserException from .login_required import LoginRequiredException from .not_found import NotFoundException from .sentry_block import SentryBlockException # Получаем класс самым простым из доступных методов SRE_Pattern = re.compile('test').__class__ __all__ = ['ServerMessageRaiser'] class ServerMessageRaiser: EXCEPTION_MAP = OrderedDict([ # # WARNING: We MUST be sure to list these exception messages in an order # which guarantees that they will be properly detected without being # detected as something else! # # For example, the "challenge_required" string ALSO exists inside of # "checkpoint_challenge_required", so if we check for ChallengeRequired # problems above CheckpointRequired, then we would ALWAYS detect # checkpoints as "challenge required" since that string exists in both # of them. # # Always list all exceptions in an order that guarantees that they # cannot be misdetected as each other! The exceptions with the longest # strings, in case of similar strings, MUST be checked/listed EARLIER! # # So in that example, CheckpointRequired MUST be listed above # ChallengeRequired! # (LoginRequiredException, [ 'login_required' ]), (CheckpointRequiredException, [ 'checkpoint_required', # message 'checkpoint_challenge_required', # error_type ]), (ChallengeRequiredException, [ 'challenge_required' ]), (FeedbackRequiredException, [ 'feedback_required' ]), (ConsentRequiredException, [ 'consent_required' ]), (IncorrectPasswordException, [ # "The password you entered is incorrect". re.compile(r'password(.+?)incorrect', re.I | re.U), # message 'bad_password', # error_type ]), (InvalidSmsCodeException, [ # "Please check the security code we sent you and try again". re.compile(r'check(.+?)security(.+?)code', re.I | re.U), # message 'sms_code_validation_code_invalid', # error_type ]), (AccountDisabledException, [ # "Your account has been disabled for violating our terms". re.compile(r'account(.+?)disabled(.+?)violating', re.I | re.U), ]), (SentryBlockException, [ 'sentry_block', ]), (InvalidUserException, [ # "The username you entered doesn't appear to belong to an account" re.compile(r'username(.+?)doesn\'t(.+?)belong', re.I | re.U), # message 'invalid_user', # error_type ]), (ForcedPasswordResetException, [ re.compile(r'reset(.+?)password', re.I | re.U), ]), ]) @classmethod def auto_raise(cls, prefix_string: Optional[str], server_message: str, api_response: ApiResponse = None, http_response: ServerResponse = None): messages = [server_message] server_error_type = None if isinstance(api_response, ApiResponse): if api_response.error_type is not None: server_error_type = api_response.error_type messages.append(server_error_type) exception_class = None def check_patterns(message, patterns): for pattern in patterns: if isinstance(pattern, SRE_Pattern): if pattern.search(message): return True elif isinstance(pattern, str): if pattern in message: return True else: raise ValueError(f'Unknown pattern type: {type(pattern)}, {pattern}') return False def iter_exceptions(message): for exception, patterns in cls.EXCEPTION_MAP.items(): if check_patterns(message, patterns): return exception for msg in messages: exception_class = iter_exceptions(msg) if exception_class is not None: break if exception_class is None: http_status_code = http_response.status if http_response is not None else None if http_status_code == 400: exception_class = BadRequestException elif http_status_code == 404: exception_class = NotFoundException else: exception_class = EndpointException display_message = server_message if server_message else server_error_type if display_message is None: display_message = 'Request failed.' message_text = f'{prefix_string}: {display_message}' if prefix_string else display_message message_text = cls.prettify_message(message_text) args = [message_text] kwargs = {} if isinstance(api_response, ApiResponse) and issubclass(exception_class, InstagramException): kwargs.update({ 'response': api_response, }) raise exception_class(*args, **kwargs) @staticmethod def prettify_message(message: str): last_char = message[-1] if message else '' if last_char not in ['', '.', '!', '?']: message += '.' message = message.capitalize() message = message.replace('_', ' ') return message
11492073
import pytest import fuzz_lightyear from fuzz_lightyear.exceptions import ConflictingHandlers from fuzz_lightyear.supplements.abstraction import get_abstraction class TestMakeRequest: def test_basic(self): def request_handler(operation_id): assert operation_id == 'status' fuzz_lightyear.make_request(request_handler) get_abstraction().request_method( 'status', ) def test_passes_other_arguments(self): def request_handler(operation_id, *args, **kwargs): assert kwargs['headers']['session'] == 'id_token' fuzz_lightyear.make_request(request_handler) get_abstraction().request_method( 'status', headers={ 'session': 'id_token', }, ) def test_throws_error_if_multiple_handlers(self): def request_handler_one(*args): pass def request_handler_two(*args): pass fuzz_lightyear.make_request(request_handler_one) with pytest.raises(ConflictingHandlers): fuzz_lightyear.make_request(request_handler_two) def test_custom_swagger_client(): def declaration(): return 1 fuzz_lightyear.custom_swagger_client(declaration) assert get_abstraction().client == 1
11492159
import numpy as np import pandas as pd from sfrmaker.routing import find_path, make_graph def valid_rnos(rnos): """Check that unique reach numbers (rno in MODFLOW 6) are consecutive and start at 1. """ sorted_reaches = sorted(rnos) consecutive = np.diff(sorted_reaches).sum() \ == len(rnos) - 1 onebased = np.min(sorted_reaches) == 1 return consecutive & onebased def valid_nsegs(nsegs, outsegs=None, increasing=True): """Check that segment numbers are valid. Parameters ---------- nsegs : list of segment numbers outsegs : list of corresponding routing connections Required if increasing=True. increasing : bool If True, segment numbers must also only increase downstream. """ # cast to array if list or series nsegs = np.atleast_1d(nsegs) outsegs = np.atleast_1d(outsegs) consecutive_and_onebased = valid_rnos(nsegs) if increasing: assert outsegs is not None graph = make_graph(nsegs, outsegs, one_to_many=False) monotonic = [] for s in nsegs: try: seg_sequence = find_path(graph.copy(), s)[:-1] # last number is 0 for outlet except: j=2 monotonic.append(np.all(np.diff(np.array(seg_sequence)) > 0)) monotonic = np.all(monotonic) return consecutive_and_onebased & monotonic else: return consecutive_and_onebased def is_to_one(toid_sequence): if isinstance(toid_sequence, dict): toid_sequence = list(toid_sequence.values()) squeezed = np.squeeze(list(toid_sequence)) if squeezed.ndim == 0: return True else: return np.isscalar(np.squeeze(list(toid_sequence))[0]) def rno_nseg_routing_consistent(nseg, outseg, iseg, ireach, rno, outreach): """Check that routing of segments (MODFLOW-2005 style) is consistent with routing between unique reach numbers (rno; MODFLOW 6 style) Parameters ---------- nseg : list or 1D numpy array outseg : list or 1D numpy array iseg : list or 1D numpy array ireach : list or 1D numpy array rno : list or 1D numpy array outreach : list or 1D numpy array Returns ------- consistent : bool """ df = pd.DataFrame({'nseg': nseg, 'outseg': outseg, 'iseg': iseg, 'ireach': ireach, 'rno': rno, 'outreach': outreach}) df.sort_values(by=['iseg', 'ireach'], inplace=True) segment_routing = dict(zip(nseg, outseg)) seg_groups = df.groupby('iseg') # segments associated with reach numbers that are first reaches first_reaches = seg_groups.first() rno1_segments = dict(zip(first_reaches.rno, first_reaches.index)) segments_consistent = [] for s, g in seg_groups: # since the segment is sorted, # rno[i+1] should == outreach[i] if len(g) > 1: preceding_consistent = np.array_equal(g.rno.values[1:], g.outreach.values[:-1]) # segments of length 1 don't have any internal routing to check else: preceding_consistent = True # check that last reach goes to same segment # as outseg in segment_routing last_outreach = g.outreach.iloc[-1] next_segment = rno1_segments.get(last_outreach, 0) last_consistent = next_segment == segment_routing[s] if not preceding_consistent & last_consistent: j=2 segments_consistent.append(preceding_consistent & last_consistent) return np.all(segments_consistent) def routing_numbering_is_valid(nseg, outseg, iseg, ireach, rno, outreach, increasing_nseg=True): """Check that routing numbering for an SFR dataset is valid. * verify that segment numbering is consecutive and starts at 1 * optionally verify that segment number only increase downstream * verify that unique reach numbering (e.g. rno in MODFLOW 6) is consecutive and starts at 1 * check that routing is consistent between segment connections (MODFLOW-2005 convention of nseg -> outseg) and reach connections (MODFLOW 6 convention based on rno) An additional check would be all non-outlet connections are listed in nseg and/or rno, but these can be assumed to be outlets (converted to 0) without modifying the nseg or rno. Parameters ---------- nseg : list or 1D numpy array outseg : list or 1D numpy array iseg : list or 1D numpy array ireach : list or 1D numpy array rno : list or 1D numpy array outreach : list or 1D numpy array increasing_nseg : bool If True, segment numbers must also only increase downstream. Returns ------- valid """ return valid_rnos(rno) & \ valid_nsegs(nseg, outseg, increasing=increasing_nseg) & \ rno_nseg_routing_consistent(nseg, outseg, iseg, ireach, rno, outreach) def routing_is_circular(fromid, toid): """Verify that segments or reaches never route to themselves. Parameters ---------- fromid : list or 1D array e.g. COMIDS, segments, or rnos toid : list or 1D array routing connections """ fromid = np.atleast_1d(fromid) toid = np.atleast_1d(toid) graph = make_graph(fromid, toid, one_to_many=False) paths = {fid: find_path(graph, fid) for fid in graph.keys()} # a fromid should not appear more than once in its sequence for k, v in paths.items(): if v.count(k) > 1: return True return False def same_sfr_numbering(reach_data1, reach_data2): """Compare two sets of reach data. Parameters ---------- reach_data1 : DataFrame Must have columns: i : zero-based row j : zero-based column iseg : segment number ireach : reach number reach_data2 : DataFrame Must have same columns as reach_data1 Returns ------- issame : bool Whether the two datasets have the same numbering for i, j andn iseg/ireach. Notes ----- k (layer) is not tested because k can be different for the same SFR package depending on the context. For example, a reach might have k=1 in the input file, and k=3 in the output file if the flux was placed in the highest active layer. """ cols = ['i', 'j', 'iseg', 'ireach'] rd1 = reach_data1[cols].sort_values(by=['iseg', 'ireach']).copy() rd2 = reach_data2[cols].sort_values(by=['iseg', 'ireach']).copy() col_equal = [] for c in rd1.columns: col_equal.append(np.array_equal(rd1[c], rd2[c])) return np.all(col_equal) def reach_elevations_decrease_downstream(reach_data): """Verify that reach values decrease monotonically in the downstream direction.""" rd = reach_data.reset_index() return check_monotonicity(rd.rno, rd.outreach, rd.strtop) #elev = dict(zip(rd.rno, rd.strtop)) #dnelev = {rid: elev[rd.outreach[i]] if rd.outreach[i] != 0 #else -9999 for i, rid in enumerate(rd.rno)} #diffs = np.array([(dnelev[i] - elev[i]) if dnelev[i] != -9999 # else -.001 for i in rd.rno]) #return np.max(diffs) <= 0 def check_monotonicity(ids, toids, values, decrease=True): """Verify that values decrease or increase monotonically in the downstream direction. Parameters ---------- ids : sequence Sequence of line identifiers (e.g. COMIDs) toids : sequence Sequence of downstream routing connections (line identifiers) values : numeric Values to check. decrease : bool If True, verify that values strictly decrease in the downstream direction, if False, verify that values strictly increase in the downstream direction. Returns ------- is_monotonic : bool Whether or not values change monotonically in the downstream direction. """ if isinstance(ids, pd.Series): ids = ids.values if isinstance(toids, pd.Series): toids = toids.values # a fill value that is larger than any value in values # (in absolute terms) default = -10 * np.max(np.abs(values)) values = np.array(values) if not decrease: values *= -1 values_dict = dict(zip(ids, values)) downstream_values = {rid: values_dict[toids[i]] if toids[i] != 0 else default for i, rid in enumerate(ids)} diffs = np.array([(downstream_values[i] - values_dict[i]) if downstream_values[i] != default else -.001 for i in ids]) return np.max(diffs) <= 0
11492202
import copy import json import re import numpy as np import torch import transformers.data.processors.squad as squad from tqdm import tqdm class SquadExample: """ A single training/test example for the Squad dataset, as loaded from disk. Args: qas_id: The example's unique identifier question_text: The question string context_text: The context string answer_text: The answer string start_position_character: The character position of the start of the answer title: The title of the example answers: None by default, this is used during evaluation. Holds answers as well as their start positions. is_impossible: False by default, set to True if the example has no possible answer. """ def __init__( self, qas_id, question_text, context_text, answer_text, start_position_character, title, answers=[], is_impossible=False, doc_tokens=None, char_to_word_offset=None, ): self.qas_id = qas_id self.question_text = question_text self.context_text = context_text self.answer_text = answer_text self.title = title self.is_impossible = is_impossible self.answers = answers self.start_position, self.end_position = 0, 0 if doc_tokens is None: doc_tokens = [] char_to_word_offset = [] prev_is_whitespace = True # Split on whitespace so that different tokens may be attributed to their original position. for c in self.context_text: if squad._is_whitespace(c): prev_is_whitespace = True else: if prev_is_whitespace: doc_tokens.append(c) else: doc_tokens[-1] += c prev_is_whitespace = False char_to_word_offset.append(len(doc_tokens) - 1) self.doc_tokens = doc_tokens self.char_to_word_offset = char_to_word_offset # Start and end positions only has a value during evaluation. if start_position_character is not None and not is_impossible: self.start_position = char_to_word_offset[start_position_character] self.end_position = char_to_word_offset[ min(start_position_character + len(answer_text) - 1, len(char_to_word_offset) - 1) ] def squad_convert_example_to_features( example, tokenizer, max_seq_length, doc_stride, max_query_length, padding_strategy, is_training, tok_to_orig_index=None, orig_to_tok_index=None, all_doc_tokens=None, ): features = [] if is_training and not example.is_impossible: # Get start and end position start_position = example.start_position end_position = example.end_position # If the answer cannot be found in the text, then skip this example. actual_text = " ".join(example.doc_tokens[start_position : (end_position + 1)]) cleaned_answer_text = " ".join(squad.whitespace_tokenize(example.answer_text)) if actual_text.find(cleaned_answer_text) == -1: return [], None, None, None if tok_to_orig_index is None: tok_to_orig_index = [] orig_to_tok_index = [] all_doc_tokens = [] for (i, token) in enumerate(example.doc_tokens): orig_to_tok_index.append(len(all_doc_tokens)) if tokenizer.__class__.__name__ in [ "RobertaTokenizer", "LongformerTokenizer", "BartTokenizer", "RobertaTokenizerFast", "LongformerTokenizerFast", "BartTokenizerFast", ]: sub_tokens = tokenizer.tokenize(token, add_prefix_space=True) else: sub_tokens = tokenizer.tokenize(token) for sub_token in sub_tokens: tok_to_orig_index.append(i) all_doc_tokens.append(sub_token) if is_training and not example.is_impossible: tok_start_position = orig_to_tok_index[example.start_position] if example.end_position < len(example.doc_tokens) - 1: tok_end_position = orig_to_tok_index[example.end_position + 1] - 1 else: tok_end_position = len(all_doc_tokens) - 1 (tok_start_position, tok_end_position) = squad._improve_answer_span( all_doc_tokens, tok_start_position, tok_end_position, tokenizer, example.answer_text ) spans = [] truncated_query = tokenizer.encode( example.question_text, add_special_tokens=False, truncation=True, max_length=max_query_length ) # Tokenizers who insert 2 SEP tokens in-between <context> & <question> need to have special handling # in the way they compute mask of added tokens. tokenizer_type = type(tokenizer).__name__.replace("Tokenizer", "").lower() sequence_added_tokens = ( tokenizer.model_max_length - tokenizer.max_len_single_sentence + 1 if tokenizer_type in squad.MULTI_SEP_TOKENS_TOKENIZERS_SET else tokenizer.model_max_length - tokenizer.max_len_single_sentence ) sequence_pair_added_tokens = tokenizer.model_max_length - tokenizer.max_len_sentences_pair span_doc_tokens = all_doc_tokens while len(spans) * doc_stride < len(all_doc_tokens): # Define the side we want to truncate / pad and the text/pair sorting if tokenizer.padding_side == "right": texts = truncated_query pairs = span_doc_tokens truncation = squad.TruncationStrategy.ONLY_SECOND.value else: texts = span_doc_tokens pairs = truncated_query truncation = squad.TruncationStrategy.ONLY_FIRST.value encoded_dict = tokenizer.encode_plus( # TODO(thom) update this logic texts, pairs, truncation=truncation, padding=padding_strategy, max_length=max_seq_length, return_overflowing_tokens=True, stride=max_seq_length - doc_stride - len(truncated_query) - sequence_pair_added_tokens, return_token_type_ids=True, ) paragraph_len = min( len(all_doc_tokens) - len(spans) * doc_stride, max_seq_length - len(truncated_query) - sequence_pair_added_tokens, ) if tokenizer.pad_token_id in encoded_dict["input_ids"]: if tokenizer.padding_side == "right": non_padded_ids = encoded_dict["input_ids"][: encoded_dict["input_ids"].index(tokenizer.pad_token_id)] else: last_padding_id_position = ( len(encoded_dict["input_ids"]) - 1 - encoded_dict["input_ids"][::-1].index(tokenizer.pad_token_id) ) non_padded_ids = encoded_dict["input_ids"][last_padding_id_position + 1 :] else: non_padded_ids = encoded_dict["input_ids"] tokens = tokenizer.convert_ids_to_tokens(non_padded_ids) token_to_orig_map = {} for i in range(paragraph_len): index = len(truncated_query) + sequence_added_tokens + i if tokenizer.padding_side == "right" else i token_to_orig_map[index] = tok_to_orig_index[len(spans) * doc_stride + i] encoded_dict["paragraph_len"] = paragraph_len encoded_dict["tokens"] = tokens encoded_dict["token_to_orig_map"] = token_to_orig_map encoded_dict["truncated_query_with_special_tokens_length"] = len(truncated_query) + sequence_added_tokens encoded_dict["token_is_max_context"] = {} encoded_dict["start"] = len(spans) * doc_stride encoded_dict["length"] = paragraph_len spans.append(encoded_dict) if "overflowing_tokens" not in encoded_dict or ( "overflowing_tokens" in encoded_dict and len(encoded_dict["overflowing_tokens"]) == 0 ): break span_doc_tokens = encoded_dict["overflowing_tokens"] for doc_span_index in range(len(spans)): for j in range(spans[doc_span_index]["paragraph_len"]): is_max_context = squad._new_check_is_max_context(spans, doc_span_index, doc_span_index * doc_stride + j) index = ( j if tokenizer.padding_side == "left" else spans[doc_span_index]["truncated_query_with_special_tokens_length"] + j ) spans[doc_span_index]["token_is_max_context"][index] = is_max_context for span in spans: # Identify the position of the CLS token cls_index = span["input_ids"].index(tokenizer.cls_token_id) p_mask = np.array([]) # # p_mask: mask with 1 for token than cannot be in the answer (0 for token which can be in an answer) # # Original TF implementation also keep the classification token (set to 0) # p_mask = np.ones_like(span["token_type_ids"]) # if tokenizer.padding_side == "right": # p_mask[len(truncated_query) + sequence_added_tokens :] = 0 # else: # p_mask[-len(span["tokens"]) : -(len(truncated_query) + sequence_added_tokens)] = 0 # pad_token_indices = np.where(span["input_ids"] == tokenizer.pad_token_id) # special_token_indices = np.asarray( # tokenizer.get_special_tokens_mask(span["input_ids"], already_has_special_tokens=True) # ).nonzero() # p_mask[pad_token_indices] = 1 # p_mask[special_token_indices] = 1 # # Set the cls index to 0: the CLS index can be used for impossible answers # p_mask[cls_index] = 0 span_is_impossible = example.is_impossible start_position = 0 end_position = 0 if is_training and not span_is_impossible: # For training, if our document chunk does not contain an annotation # we throw it out, since there is nothing to predict. doc_start = span["start"] doc_end = span["start"] + span["length"] - 1 out_of_span = False if not (tok_start_position >= doc_start and tok_end_position <= doc_end): out_of_span = True if out_of_span: start_position = cls_index end_position = cls_index span_is_impossible = True else: if tokenizer.padding_side == "left": doc_offset = 0 else: doc_offset = len(truncated_query) + sequence_added_tokens start_position = tok_start_position - doc_start + doc_offset end_position = tok_end_position - doc_start + doc_offset features.append( squad.SquadFeatures( span["input_ids"], span["attention_mask"], span["token_type_ids"], cls_index, p_mask.tolist(), example_index=0, # Can not set unique_id and example_index here. They will be set after multiple processing. unique_id=0, paragraph_len=span["paragraph_len"], token_is_max_context=span["token_is_max_context"], tokens=span["tokens"], token_to_orig_map=span["token_to_orig_map"], start_position=start_position, end_position=end_position, is_impossible=span_is_impossible, qas_id=example.qas_id, ) ) return features, tok_to_orig_index, orig_to_tok_index, all_doc_tokens def sample_writer(data, config, tokenizer, is_train): """process and write single 'paragraph-context' from squad-formed QA dataset""" context = data["context"] example = None tok_to_orig_index = None orig_to_tok_index = None all_doc_tokens = None write_data = [] for qas in data["qas"]: is_impossible = qas.get("is_impossible", False) example = SquadExample( qas_id=qas["id"], question_text=qas["question"], context_text=context, answer_text="" if is_impossible else qas["answers"][0]["text"], start_position_character=0 if is_impossible else qas["answers"][0]["answer_start"], title="", answers=qas["answers"], is_impossible=is_impossible, doc_tokens=(None if example is None else example.doc_tokens), char_to_word_offset=(None if example is None else example.char_to_word_offset), ) features, tok_to_orig_index, orig_to_tok_index, all_doc_tokens = squad_convert_example_to_features( example=example, tokenizer=tokenizer, max_seq_length=config.max_seq_length, doc_stride=config.doc_stride, max_query_length=config.max_query_length, padding_strategy="max_length", is_training=is_train, tok_to_orig_index=tok_to_orig_index, orig_to_tok_index=orig_to_tok_index, all_doc_tokens=all_doc_tokens, ) for i, feature in enumerate(features): write_datum = { "input_ids": feature.input_ids, "attention_mask": feature.attention_mask, "token_type_ids": feature.token_type_ids, "start_position": feature.start_position, "end_position": feature.end_position, } if not is_train: write_datum["id"] = feature.qas_id write_datum["unique_id"] = "{}_{}".format(feature.qas_id, i) write_datum["tokens"] = feature.tokens write_datum["token_to_orig_map"] = feature.token_to_orig_map write_datum["paragraph_len"] = feature.paragraph_len write_datum["token_is_max_context"] = feature.token_is_max_context if i == 0: # only store example data for the first feature from the example write_datum["is_impossible"] = example.is_impossible write_datum["answers"] = example.answers write_datum["doc_tokens"] = example.doc_tokens else: write_datum["is_impossible"] = None write_datum["answers"] = None write_datum["doc_tokens"] = None write_data.append(write_datum) return write_data def process_korquad_1(config, data_file, train): with open(data_file) as handle: load_data = json.load(handle)["data"] data = [] for datum in load_data: data.extend(datum["paragraphs"]) return data def process_korquad_2(config, data_file, train): TAGS = [ "<!DOCTYPE html>", "<html>", "</html>", "<meta>", "<head>", "</head>", "<body>", "</body>", "<label>", "</label>", "<div>", "</div>", "<a>", "</a>", "<span>", "</span>", "<link>", "<img>", "<sup>", "</sup>", "<input>", "<noscript></noscript>", "<p>", "</p>", "<cite>", "</cite>", "<tbody>", "</tbody>", "</br>", "<br/>", "<h1>", "</h1>", "<h2>", "</h2>", "<h3>", "</h3>", "<form>", "</form>", "<small>", "</small>", "<big>", "</big>", "<b>", "</b>", "<abbr>", "</abbr>", "[편집]", ] def remove_tags(text): # only retain meaningful html tags to extract answers # tags like <table> <tr> <td> <ul> <li> will remain text = re.sub("<title>.*?</title>", " ", text) text = text.replace('rowspan="', "r") # summarize attribute name for col, row spans text = text.replace('colspan="', "c") # which indicate merged table cells for tag in TAGS: text = text.replace(tag, "") text = re.sub(" +", " ", text) text = re.sub("\n+", "\n", text) return text with open(data_file) as handle: data = json.load(handle)["data"] qas_num = [] for datum in data: context = datum["context"] datum["context"] = remove_tags(context) del datum["raw_html"] if train and not config.all_korquad_2_sample: # only use first sample from context for train split datum["qas"] = datum["qas"][:1] for qas in datum["qas"]: del qas["answer"]["html_answer_text"] del qas["answer"]["html_answer_start"] # qas["id"] = "{}-{}".format(i, qas["id"]) answer_prev_text = context[: qas["answer"]["answer_start"]] answer_prev_len = len(answer_prev_text) remove_tag_text = remove_tags(answer_prev_text) # adjust span position according to text without tags qas["answer"]["answer_start"] -= answer_prev_len - len(remove_tag_text) qas["answer"]["text"] = remove_tags(qas["answer"]["text"]) qas["answers"] = [qas.pop("answer")] qas_num.append(len(datum["qas"])) # some paragraph in korquad_2 have too many question samples # it will slow down the progress of multiprocessing job # so, i split them limit_qas_num = int(np.percentile(np.array(qas_num), 50)) flat_data = [] for datum in data: qas_num = len(datum["qas"]) if qas_num > limit_qas_num: for j in range(0, qas_num, limit_qas_num): _datum = copy.deepcopy(datum) _datum["qas"] = datum["qas"][j : j + limit_qas_num] flat_data.append(_datum) else: flat_data.append(datum) data = flat_data return data def process_kluemrc(config, data_file, train): data = [] with open(data_file) as handle: for datum in json.load(handle)["data"]: datum = datum["paragraphs"] for qas in datum: for q in qas["qas"]: q["id"] = q.pop("guid") data.extend(datum) return data def process_tydiqa(config, data_file, train): data = [] total = sum([1 for _ in open(data_file)]) for line in tqdm(open(data_file), total=total, dynamic_ncols=True): datum = json.loads(line) if datum["language"].lower().strip() != "korean": continue span_byte_map = {} prev_bytes = 0 for i, char in enumerate(datum["document_plaintext"]): byte_len = len(char.encode("utf-8")) for j in range(byte_len): span_byte_map[prev_bytes + j] = i prev_bytes += byte_len answers = [] bool_answers = [] is_impossible = False for annot in datum["annotations"]: spans = annot["minimal_answer"] start = spans["plaintext_start_byte"] end = spans["plaintext_end_byte"] yesno = None if annot["yes_no_answer"] == "NONE" else annot["yes_no_answer"] if yesno is not None: bool_answers.append(yesno) continue if spans["plaintext_start_byte"] == -1: is_impossible = True else: start = span_byte_map[start] end = span_byte_map[end] answers.append( { "text": datum["document_plaintext"][start:end], "answer_start": start, } ) # skip boolqa samples if len(bool_answers) != 0: continue if len(answers) != 0: is_impossible = False else: is_impossible = True data.append( { "context": datum["document_plaintext"], "qas": [ { "id": len(data), "question": datum["question_text"], "is_impossible": is_impossible, "answers": answers, } ], } ) return data process_map = { "korquad_1": process_korquad_1, "korquad_2": process_korquad_2, "tydiqa": process_tydiqa, "kluemrc": process_kluemrc, } def collate_fn(features): input_ids = [sample["input_ids"] for sample in features] attention_mask = [sample["attention_mask"] for sample in features] token_type_ids = [sample["token_type_ids"] for sample in features] start_position = [sample["start_position"] for sample in features] end_position = [sample["end_position"] for sample in features] input_ids = torch.tensor(np.array(input_ids).astype(np.int64), dtype=torch.long) attention_mask = torch.tensor(np.array(attention_mask).astype(np.int8), dtype=torch.long) token_type_ids = torch.tensor(np.array(token_type_ids).astype(np.int8), dtype=torch.long) start_position = torch.tensor(np.array(start_position).astype(np.int64), dtype=torch.long) end_position = torch.tensor(np.array(end_position).astype(np.int64), dtype=torch.long) inputs = { "input_ids": input_ids, "attention_mask": attention_mask, "token_type_ids": token_type_ids, "start_positions": start_position, "end_positions": end_position, } if "unique_id" in features[0]: inputs["unique_id"] = [sample["unique_id"] for sample in features] return inputs
11492204
import pytest from typus import en_typus, ru_typus from typus.chars import * QUOTES = ( ''.join((LAQUO, RAQUO, DLQUO, LDQUO)), ''.join((LDQUO, RDQUO, LSQUO, RSQUO)), ) TYPUSES = ( (ru_typus, {}), (en_typus, str.maketrans(*QUOTES)), ) @pytest.fixture(name='assert_typus', scope='module', params=TYPUSES) def get_assert_typus(request): typus, charmap = request.param def assert_typus(source, expected): assert typus(source) == expected.translate(charmap) return assert_typus def test_debug(): assert ru_typus('1m', debug=True) == '1_m' @pytest.mark.parametrize('source, expected', ( ('00 "11" 00', '00 «11» 00'), # clashes with digit_spaces ( '''00" "11 '22' 11"? "11 '22 "33 33?"' 11" 00 "11 '22' 11" 0"''', f'00{DPRIME} «11 „22“ 11»? «11 „22 «33{NBSP}33?»“ 11» ' f'00 «11 „22“ 11» 0{DPRIME}' ), )) def test_quotes(assert_typus, source, expected): assert_typus(source, expected) @pytest.mark.parametrize('source, expected', ( ('--', '--'), ('foo - foo', f'foo{MDASH_PAIR}foo'), # Leading comma case (', - foo', f',{MDASH}{THNSP}foo'), (', -- foo', f',{MDASH}{THNSP}foo'), # if line begins, adds nbsp after mdash ('-- foo', f'{MDASH}{NBSP}foo'), # if line ends, adds nbsp before mdash ('foo --', f'foo{NBSP}{MDASH}'), ('foo -- bar', f'foo{MDASH_PAIR}bar'), # Python markdown replaces dash with ndash, don't know why (f'foo {NDASH} foo', f'foo{MDASH_PAIR}foo'), # This one for ru_typus ('foo - "11" 00', f'foo{MDASH_PAIR}«11» 00'), ('2 - 2foo', f'2{MDASH_PAIR}2foo'), # no units clash ('2 - 2', f'2{NBSP}{MINUS}{NBSP}2'), # + minus ('Winnie-the-Pooh', 'Winnie-the-Pooh'), )) def test_mdash(assert_typus, source, expected): assert_typus(source, expected) @pytest.mark.parametrize('source, expected', ( ('"4"', '«4»'), ('4\'', '4' + SPRIME), ('4"', '4' + DPRIME), ('" 22"', '" 22' + DPRIME), )) def test_primes(assert_typus, source, expected): assert_typus(source, expected) @pytest.mark.parametrize('source, expected', ( ('25-foo', '25-foo'), ('2-3', f'2{NDASH}3'), ('2,5-3', f'2,5{NDASH}3'), ('0.5-3', f'0.5{NDASH}3'), ('2-3 foo', f'2{NDASH}3{NBSP}foo'), # + ranges ('(15-20 items)', f'(15{NDASH}20{NBSP}items)'), # Float ('0,5-3', f'0,5{NDASH}3'), ('-0,5-3', f'{MINUS}0,5{NDASH}3'), ('-5.5-3', f'{MINUS}5.5{NDASH}3'), ('-5,5-3', f'{MINUS}5,5{NDASH}3'), ('-5,5-3.5', f'{MINUS}5,5{NDASH}3.5'), ('2 - 3', f'2{NBSP}{MINUS}{NBSP}3'), ('2-3 x 4', f'2{MINUS}3{NBSP}{TIMES}{NBSP}4'), ('2-3 * 4', f'2{MINUS}3{NBSP}{TIMES}{NBSP}4'), ('2-3 - 4', f'2{MINUS}3{NBSP}{MINUS}{NBSP}4'), )) def test_ranges(assert_typus, source, expected): assert_typus(source, expected) @pytest.mark.parametrize('source, expected', ( # Minus (f'3{NBSP}-{NBSP}2', f'3{NBSP}{MINUS}{NBSP}2'), # This one clashes with range ('2-3', f'2{NDASH}3'), # This one clashes with mdash (f'x{NBSP}-{NBSP}3', f'x{NNBSP}{MDASH}{THNSP}3'), ('-3', f'{MINUS}3'), # Star ('3*2', f'3{TIMES}2'), ('*3', f'{TIMES}3'), (f'3{NBSP}*{NBSP}2', f'3{NBSP}{TIMES}{NBSP}2'), (f'x{NBSP}*{NBSP}2', f'x{NBSP}{TIMES}{NBSP}2'), # 'x' ('3x2', f'3{TIMES}2'), ('x3', f'{TIMES}3'), (f'3{NBSP}x{NBSP}2', f'3{NBSP}{TIMES}{NBSP}2'), (f'x{NBSP}x{NBSP}2', f'x{NBSP}{TIMES}{NBSP}2'), # and Russian "х" ('3х2', f'3{TIMES}2'), ('х3', f'{TIMES}3'), (f'3{NBSP}х{NBSP}2', f'3{NBSP}{TIMES}{NBSP}2'), (f'x{NBSP}х{NBSP}2', f'x{NBSP}{TIMES}{NBSP}2'), )) def test_math(assert_typus, source, expected): assert_typus(source, expected) @pytest.mark.parametrize('source, expected', ( ('aaa 2a', f'aaa 2a'), # doesnt clash with units )) def test_pairs(assert_typus, source, expected): assert_typus(source, expected) @pytest.mark.parametrize('source, expected', ( ('4444444 fooo', '4444444 fooo'), ('444 foo', f'444{NBSP}foo'), ('444 +', f'444{NBSP}+'), ('444 4444 bucks', f'444{NBSP}4444 bucks'), ('4444444 foo', f'4444444 foo'), # no untis clash ('444 -', f'444{NBSP}{MDASH}'), )) def test_digit_spaces(assert_typus, source, expected): assert_typus(source, expected) def test_example(assert_typus): source = ( 'Излучение, как следует из вышесказанного, концентрирует ' 'внутримолекулярный предмет - деятельности . "...ff \'Можно?\' ' 'предположить, что силовое - "поле "мент "d" ально" отклоняет" ' 'сенсибельный \'квазар !..\' cc", не учитывая мнения авторитетов. ' 'Искусство испускает данный электрон, учитывая опасность, ' '<code> "d" test -- test(c)</code> которую представляли ' 'собой писания Дюринга для не окрепшего еще немецкого рабочего ' 'движения. Смысл жизни -- амбивалентно (с) дискредитирует ' 'закон (r) исключённого(tm) третьего (тм)... \n\n\n' '1500 мА*ч\n\n' '1-2=4\n' '- Химическое соединение (p) ненаблюдаемо контролирует экран-ый ' 'квазар (р). Идеи 3/4 гедонизма занимают b & b центральное место ' 'в утилитаризме(sm) "<NAME>", однако <- гравитирующая -> ' 'сфера масштабирует фотон, +-2мм изменяя привычную == реальность. ' 'Силовое *3 поле -3 реально 3 * 2 /= 6 3x3 восстанавливает ' 'трансцендентальный 3" 2\' принцип 1000р. восприятия.' '"...\'test\'" (c) m&m\'s\n\n\n' ) expected = ( 'Излучение, как следует из_вышесказанного, концентрирует ' 'внутримолекулярный предмет\u202f—\u2009деятельности. «…ff „Можно?“ ' 'предположить, что силовое\u202f—\u2009„поле «мент „d“ ально» ' 'отклоняет“ ' 'сенсибельный „квазар!..“ cc», не_учитывая мнения авторитетов. ' 'Искусство испускает данный электрон, учитывая опасность, ' '<code> "d" test -- test(c)</code> которую представляли собой ' 'писания Дюринга для не_окрепшего еще немецкого рабочего ' 'движения. Смысл жизни\u202f—\u2009амбивалентно ©_дискредитирует ' 'закон® исключённого™ третьего™…\n\n' '1500_мА•ч\n\n' '1−2=4\n' '—_Химическое соединение℗ ненаблюдаемо контролирует экран-ый ' 'квазар℗. Идеи ¾_гедонизма занимают b_&_b_центральное место ' 'в_утилитаризме℠ «Милля и_Бентама», однако ←_гравитирующая_→ ' 'сфера масштабирует фотон, ±2_мм изменяя привычную_≡_реальность. ' 'Силовое ×3_поле −3_реально 3_×_2_≠_6 3×3 восстанавливает ' 'трансцендентальный 3″ 2′ принцип 1000_₽ восприятия.' '«…„test“» ©_m&m’s' ).replace('_', NBSP) assert_typus(source, expected)
11492211
from pandas_confusion import ConfusionMatrix def write_prediction_file(path, test_unique_ids, Y_test_indices, y_prediction_classes): with open(path, 'w') as prediction_handler: prediction_handler.write('{}\t{}\t{}\n'.format("UniqueID", "Gold_Label", "Prediction")) for index, val in enumerate(test_unique_ids): prediction_handler.write('{}\t{}\t{}\n'.format(val, Y_test_indices[index], y_prediction_classes[index])) def write_score_file(score_file, f1_mean, f1, model, Y_test_indices, y_prediction_classes): with open(score_file, 'w') as score_handler: score_handler.write("Micro-averaged F1: {}\n".format(f1_mean)) score_handler.write("Individual scores: {}\n".format(f1)) score_handler.write("Confusion matrix:\n") score_handler.write(str(ConfusionMatrix(Y_test_indices, y_prediction_classes))) score_handler.write("\n\n\nModel summary\n") model.summary(print_fn=lambda x: score_handler.write(x + '\n'))
11492227
import numpy as np from bokeh.core.properties import Float from bokeh.io import save from bokeh.model import DataModel from bokeh.models import ColumnDataSource, CustomJS from bokeh.plotting import figure class Params(DataModel): amp = Float(default=0.1, help="Amplitude") freq = Float(default=0.1, help="Frequency") phase = Float(default=0, help="Phase") offset = Float(default=-5, help="Offset") params = Params(amp=2, freq=3, phase=0.4, offset=1) A = params.amp k = params.freq phi = params.phase B = params.offset x = np.linspace(0, 10, 100) y = A*np.sin(k*x + phi) + B source = ColumnDataSource(data=dict(x=x, y=y)) plot = figure(tags=[params], y_range=(-10, 10), title="Data models example") plot.line("x", "y", source=source, line_width=3, line_alpha=0.6) callback = CustomJS(args=dict(source=source, params=params), code=""" const data = source.data const A = params.amp const k = params.freq const phi = params.phase const B = params.offset const {x, y} = data for (let i = 0; i < x.length; i++) { y[i] = A*Math.sin(k*x[i] + phi) + B } source.change.emit() """) params.js_on_change("amp", callback) params.js_on_change("freq", callback) params.js_on_change("phase", callback) params.js_on_change("offset", callback) save(plot)
11492280
import os from pathlib import Path def write_nblink(fpath, route): rpath = Path(route, fpath) new_name = fpath.name.replace(".ipynb", ".nblink") fpath = fpath.with_name(new_name) with open(fpath, "w") as f: f.write("{ \n") f.write(f' "path": "{rpath.as_posix()}"\n') f.write("}\n") dir_docs = Path("./") dir_examples_root = Path("../") dir_examples = Path(dir_examples_root, "examples") for root, subdirectories, files in os.walk(dir_examples): for file in files: fpath = Path(root, file) if ".ipynb_checkpoints" not in str(fpath) and fpath.suffix == ".ipynb": # Find path relative to examples root, e.g.: examples/filters/x.ipynb rpath = fpath.relative_to(dir_examples_root) # Create a similar directory structure under docs/ rpath.parent.mkdir(parents=True, exist_ok=True) # Generate and write .nblink file in the created directory structure num_dirs_deep = len(rpath.parent.parts) route = Path("../" * num_dirs_deep, dir_examples_root) write_nblink(rpath, route)
11492282
import re import SourceModel.SM_CaseStmt import SourceModel.SM_Class import SourceModel.SM_Constants as SMCONSTS import SourceModel.SM_Define import SourceModel.SM_Define import SourceModel.SM_Element import SourceModel.SM_Exec import SourceModel.SM_FileResource import SourceModel.SM_IfStmt import SourceModel.SM_IncludeResource import SourceModel.SM_LCOM import SourceModel.SM_Node import SourceModel.SM_PackageResource import SourceModel.SM_ServiceResource import SourceModel.SM_User from SmellDetector import Utilities class SM_File: def __init__(self, file=""): if file != "": curFile = open(file, 'rt', errors='ignore') self.fileText = curFile.read() self.resourceBodyText = self.fileText self.fileName = file curFile.close() def setText(self, text): self.fileText = text def getNoOfClassDeclarations(self): return self.countEntityDeclaration(SMCONSTS.CLASS_REGEX, "class") def getNoOfDefineDeclarations(self): return self.countEntityDeclaration(SMCONSTS.DEFINE_REGEX, "define") def getNoOfFileDeclarations(self): return self.countEntityDeclaration(SMCONSTS.FILE_REGEX, "file") def getNoOfPackageDeclarations(self): return self.countEntityDeclaration(SMCONSTS.PACKAGE_REGEX, "package") def getNoOfServiceDeclarations(self): return self.countEntityDeclaration(SMCONSTS.SERVICE_REGEX, "service") def getNoOfExecDeclarations(self): return self.countEntityDeclaration(SMCONSTS.EXEC_REGEX, "exec") def getLinesOfCode(self): counter = self.countEntityDeclaration(SMCONSTS.LOC_REGEX, "newLine") if counter > 0: return counter+1 if (len(self.fileText) > 0): return 1 return 0 def getLinesOfCodeWithoutComments(self): totalLines = self.getLinesOfCode() totalCommentsLines = self.getLinesOfComments() return totalLines - totalCommentsLines def getLinesOfComments(self): counter = self.countEntityDeclaration(SMCONSTS.COMMENT_REGEX, "newLine") return counter def countEntityDeclaration(self, regEx, entityType): compiledRE = re.compile(regEx) Utilities.myPrint("Identified " + entityType + " declarations: " + str(compiledRE.findall(self.fileText)) + \ " Size: " + str(len(compiledRE.findall(self.fileText)))) return len(compiledRE.findall(self.fileText)) def getFileResourceList(self): compiledRE = re.compile(SMCONSTS.FILE_REGEX) fileResourceList = [] for match in (compiledRE.findall(self.fileText)): fileResourceText = self.extractResourceText(match) Utilities.myPrint("Extracted file declaration: " + fileResourceText) fileResourceObj = SourceModel.SM_FileResource.SM_FileResource(fileResourceText) fileResourceList.append(fileResourceObj) return fileResourceList def extractResourceText(self, initialString): index = self.fileText.find(initialString) if index < 0: return initialString compiledRE1 = re.compile(r'\{') compiledRE2 = re.compile(r'\}') curBracketCount = len(compiledRE1.findall(initialString)) - len(compiledRE2.findall(initialString)) curIndex = index + len(initialString) + 1 if curBracketCount == 0: #This is to find the first "{" since currently there is no { which may happen in case of multi-line def found = False while curIndex < len(self.fileText) and not found: if self.fileText[curIndex] == '{': found = True curBracketCount = 1 curIndex += 1 while curBracketCount > 0 and curIndex < len(self.fileText): if self.fileText[curIndex] == '}': curBracketCount -= 1 if self.fileText[curIndex] == '{': curBracketCount += 1 curIndex +=1 return self.fileText[index:curIndex] def getServiceResourceList(self): compiledRE = re.compile(SMCONSTS.SERVICE_REGEX) serviceResourceList = [] for match in (compiledRE.findall(self.fileText)): serviceResourceText = self.extractResourceText(match) Utilities.myPrint("Extracted service declaration: " + serviceResourceText) serviceResourceObj = SourceModel.SM_ServiceResource.SM_ServiceResource(serviceResourceText) serviceResourceList.append(serviceResourceObj) return serviceResourceList def getPackageResourceList(self): compiledRE = re.compile(SMCONSTS.PACKAGE_REGEX) packageResourceList = [] for match in (compiledRE.findall(self.fileText)): packageResourceText = self.extractResourceText(match) Utilities.myPrint("Extracted package declaration: " + packageResourceText) packageResourceObj = SourceModel.SM_PackageResource.SM_PackageResource(packageResourceText) packageResourceList.append(packageResourceObj) return packageResourceList def getClassDeclarationList(self): compiledRE = re.compile(SMCONSTS.CLASS_REGEX) compiledClassNameRE = re.compile(SMCONSTS.CLASS_NAME_REGEX) classList = [] for match in compiledRE.findall(self.fileText): className = compiledClassNameRE.findall(match)[0] #print("Class name: %s" % (className)) classText = self.extractResourceText(match) Utilities.myPrint("Extracted class declaration: " + classText) classObj = SourceModel.SM_Class.SM_Class(classText, className) classList.append(classObj) return classList def getDefineDeclarationList(self): compiledRE = re.compile(SMCONSTS.DEFINE_REGEX) defineList = [] for match in compiledRE.findall(self.fileText): defineText, s, e = self.extractElementText(match) Utilities.myPrint("Extracted define declaration: " + defineText) defineObj = SourceModel.SM_Define.SM_Define(defineText) defineList.append(defineObj) return defineList def getLCOM(self): return SourceModel.SM_LCOM.getLCOM(self.getOuterElementList()) def getBodyTextSize(self): loc = self.getLinesOfCode() return loc, len(self.resourceBodyText) def getOuterClassList(self): outerElementList = self.getOuterElementList() classList = [] for element in outerElementList: if type(element) is SourceModel.SM_Class.SM_Class: classList.append(element) return classList def getOuterDefineList(self): outerElementList = self.getOuterElementList() defineList = [] for element in outerElementList: if type(element) is SourceModel.SM_Define.SM_Define: defineList.append(element) return defineList # exElementList = [] # exElementList.extend(self.getElementList(SMCONSTS.DEFINE_REGEX)) # filteredList = self.filterOutInnerElements(exElementList) # return filteredList def getOuterElementList(self): exElementList = [] exElementList.extend(self.getElementList(SMCONSTS.CLASS_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.SERVICE_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.CASE_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.DEFINE_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.EXEC_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.FILE_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.IF_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.PACKAGE_REGEX)) exElementList.extend(self.getElementList(SMCONSTS.USER_REGEX)) filteredList = self.filterOutInnerElements(exElementList) return filteredList def getElementList(self, regex): compiledRE = re.compile(regex) exElementList = [] for str in (compiledRE.findall(self.fileText)): elementText, startIndex, endIndex = self.extractElementText(str) elementObj = self.getElementObject(elementText, regex) exElementList.append(ExElement(elementObj, startIndex, endIndex)) return exElementList # TODO: Handle variables # Unwrap classes from list def getIncludeClasses(self): compiledIncludeRE = re.compile(SMCONSTS.DECLARE_INCLUDE_REGEX) compiledResourceRE = re.compile(SMCONSTS.DECLARE_RESOURCE_REGEX) declareClassList = [] declareClassName = "" for match in (compiledIncludeRE.findall(self.fileText)): #print(match) declareClassText = match cleanInclude = re.sub(r'^\s*include \[?(.+)\]?\s*$', r'\1', declareClassText) #print("Clean include: %s" % cleanInclude) class_name = r'(?:Class\[)?\'?\:{0,2}([\w\d\:\-_\$]+)\'?\]?' classRE = re.compile(class_name) if ',' in cleanInclude: classes = cleanInclude.split(',') for c in classes: for m in classRE.findall(c): # Find a variable value in text if m.startswith('$'): #print("Variable: %s" % m) varRE = r'(?:^|\n)\s*\$[\w\d\-_]+\s?=\s?\'?\"?([\w\d\-_]+)\'?\"?\n' compiledVarRE = re.compile(varRE) for v in (compiledVarRE.findall(self.fileText)): #print(v) declareClassName = v Utilities.myPrint("Extracted include class declaration: " + declareClassText) declareResourceObj = SourceModel.SM_IncludeResource.SM_IncludeResource(declareClassText, declareClassName) declareClassList.append(declareResourceObj) break #print("Variable %s value) #print("Extracted class name: %s" % m) else: declareClassName = m Utilities.myPrint("Extracted include class declaration: " + declareClassText) declareResourceObj = SourceModel.SM_IncludeResource.SM_IncludeResource(declareClassText, declareClassName) declareClassList.append(declareResourceObj) else: for c in classRE.findall(cleanInclude): #print("Extracted class name: %s" % c) declareClassName = c #print("%s" % includeClassText) Utilities.myPrint("Extracted include class declaration: " + declareClassText) declareResourceObj = SourceModel.SM_IncludeResource.SM_IncludeResource(declareClassText, declareClassName) declareClassList.append(declareResourceObj) for match in (compiledResourceRE.findall(self.fileText)): #print(match) declareClassText = match declareClassName = declareClassText #print("%s" % includeClassText) Utilities.myPrint("Extracted resource class declaration: " + declareClassText) declareResourceObj = SourceModel.SM_IncludeResource.SM_IncludeResource(declareClassText, declareClassName) declareClassList.append(declareResourceObj) return declareClassList def extractElementText(self, initialString): compiledRE1 = re.compile(r'\{') compiledRE2 = re.compile(r'\}') curBracketCount = len(compiledRE1.findall(initialString)) - len(compiledRE2.findall(initialString)) index = self.fileText.find(initialString) if index < 0: return initialString, 0, len(initialString) curIndex = index + len(initialString) + 1 if curBracketCount == 0: #And now we need to find the corresponding ')' to avoid any errors where curly brackets are matched #in the parameters itself. found = False while curIndex < len(self.fileText) and not found: if self.fileText[curIndex] == ')': found = True curIndex +=1 #This is to find the first "{" since currently there is no { which may happen in case of multi-line class def found = False while curIndex < len(self.fileText) and not found: if self.fileText[curIndex] == '{': found = True curBracketCount = 1 curIndex += 1 while curBracketCount > 0 and curIndex < len(self.fileText): if self.fileText[curIndex] == '}': curBracketCount -= 1 if self.fileText[curIndex] == '{': curBracketCount += 1 curIndex +=1 return self.fileText[index:curIndex], index, curIndex def getElementObject(self, elementText, regex): if regex == SMCONSTS.CLASS_REGEX: return SourceModel.SM_Class.SM_Class(elementText) if regex == SMCONSTS.DEFINE_REGEX: return SourceModel.SM_Define.SM_Define(elementText) if regex == SMCONSTS.EXEC_REGEX: return SourceModel.SM_Exec.SM_Exec(elementText) if regex == SMCONSTS.FILE_REGEX: return SourceModel.SM_FileResource.SM_FileResource(elementText) if regex == SMCONSTS.PACKAGE_REGEX: return SourceModel.SM_PackageResource.SM_PackageResource(elementText) if regex == SMCONSTS.SERVICE_REGEX: return SourceModel.SM_ServiceResource.SM_ServiceResource(elementText) if regex == SMCONSTS.DECLARE_INCLUDE_REGEX or regex == SMCONSTS.DECLARE_RESOURCE_REGEX: return SourceModel.SM_IncludeResource.SM_IncludeResource(elementText) if regex == SMCONSTS.IF_REGEX: return SourceModel.SM_IfStmt.SM_IfStmt(elementText) if regex == SMCONSTS.CASE_REGEX: return SourceModel.SM_CaseStmt.SM_CaseStmt(elementText) if regex == SMCONSTS.USER_REGEX: return SourceModel.SM_User.SM_User(elementText) def sort(self, exClassElementList): result = [] while len(exClassElementList) > 0: largest = self.findLargest(exClassElementList) result.append(largest) exClassElementList.remove(largest) return result def findLargest(self, exClassElementList): if len(exClassElementList) > 0: largest = exClassElementList[0] for item in exClassElementList: if (item.endIndex - item.startIndex) > (largest.endIndex - item.startIndex): largest = item return largest def filterOutInnerElements(self, exClassElementList): filteredList = [] exClassElementList = self.sort(exClassElementList) for element in exClassElementList: found = False for filteredItem in filteredList: if element.startIndex >= filteredItem.startIndex and element.endIndex <= filteredItem.endIndex: found = True break if found == False: filteredList.append(element) classElementList = [] for item in filteredList: classElementList.append(item.elementObj) return classElementList def getMaxNestingDepth(self): maxNestingDepth = 0 curIndex = 0 curBracketCount = 0 while curIndex < len(self.fileText): if self.fileText[curIndex] == '}': curBracketCount -= 1 if self.fileText[curIndex] == '{': curBracketCount += 1 if curBracketCount > maxNestingDepth: maxNestingDepth = curBracketCount curIndex +=1 return maxNestingDepth def getHardCodedStatments(self): compiledRE = re.compile(SMCONSTS.HARDCODED_VALUE_REGEX) hardCodedStmtList = compiledRE.findall(self.fileText) filteredList = [] for item in hardCodedStmtList: #print(item) if not (item.__contains__("$") or item.__contains__("Package") or item.__contains__("Service") \ or item.__contains__("File")): filteredList.append(item) #print(filteredList) return filteredList def getClassHierarchyInfo(self): classDecls = self.getClassDeclarationList() classList = [] parentClassList = [] for aClass in classDecls: classes, pClasses = aClass.getClassHierarchyInfo() if len(classes) > 0: classList.append(classes) if len(pClasses) > 0: parentClassList.append(pClasses) return classList, parentClassList def getNodeDeclarations(self): compiledRE = re.compile(SMCONSTS.NODE_REGEX) nodeResourceList = [] for match in (compiledRE.findall(self.fileText)): nodeResourceText = self.extractResourceText(match) Utilities.myPrint("Extracted node declaration: " + nodeResourceText) nodeResourceObj = SourceModel.SM_Node.SM_Node(nodeResourceText) nodeResourceList.append(nodeResourceObj) return nodeResourceList class ExElement(object): def __init__(self, elementObj, startIndex, endIndex): self.elementObj = elementObj self.startIndex = startIndex self.endIndex = endIndex
11492315
import numpy as np # linear algebra import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv) import matplotlib.pyplot as plt import xgboost as xgb from xgboost import XGBClassifier, XGBRegressor from xgboost import plot_importance from catboost import CatBoostRegressor from matplotlib import pyplot import shap from time import time from tqdm import tqdm_notebook as tqdm from collections import Counter from scipy import stats import lightgbm as lgb from sklearn.metrics import cohen_kappa_score, mean_squared_error from sklearn.model_selection import KFold, StratifiedKFold import gc import json from random import choice import tensorflow as tf from sklearn.preprocessing import StandardScaler, MinMaxScaler, OneHotEncoder pd.set_option('display.max_columns', 1000) # fast cappa eval function for lgb def eval_qwk_lgb_regr(y_true, y_pred): dist = Counter(reduce_train['accuracy_group']) for k in dist: dist[k] /= len(reduce_train) reduce_train['accuracy_group'].hist() acum = 0 bound = {} for i in range(3): acun += dist[i] bound[i] = np.percentile(y_pred, acum * 100) def classify(x): if x <= bound[0]: return 0 elif x<= bound[1]: return 1 elif x <= bound[2]: return 2 else : return 3 y_pred = np.array(list(map(classify, y_pred))).reshape(y_true.shape) return 'cappa', cohen_kappa_score(y_true, y_pred, weights = 'quadratic'), True def cohenkappa(y_pred, y): y = y.get_label().astype("int") ypred = ypred.reshape((4, -1)).argmax(axis = 0) loss = cohen_kappa_score(y, y_pred, weights = 'quadradtic') return "cappa", loss, True def read_data(): print('Reading train.csv file...') train = pd.read_csv('train.csv') print('Reading test.csv file...') test = pd.read_csv('test.csv') print('Reading train_labels.csv file...') train_labels = pd.read_csv('train_labels.csv') print('Reading specs.csv file...') specs = pd.read_csv('specs.csv') print('Reading sample_submission.csv file...') sample_submission = pd.read_csv('sample_submission') return train, test, train_labels, specs, sample_submission def encode_title(train, test, train_labels): # encode title train['title_event_code'] = list(map(lambda x,y : str(x) + '_' +str(y), train['title'], train['event_code'])) test['title_event_code'] = list(map(lambda x,y : str(x) + '_' + str(y), test['title'], test['event_code'])) all_title_event_code = list(set(train['title_event_code'].unique()). union(test["title_event_code"].unique())) # make a list with all unique "titles" list_of_user_activities = list(set(train['title'].unique()).union(set(test['title'].unique()))) # make a list with all the unique 'event code' list_of_event_code = list(set(train['event_code'].unique()).union(set(test['event_code'].unique()))) list_of_event_id = list(set(train['event_id'].unique()).union(set(test['event_id'].unique()))) # make a list with all the unique worlds list_of_worlds = list(set(train['world'].unique()).union(set(test['world'].unique()))) # create a dictionat numerating the titles activities_map = dict(zip(list_of_user_activities, np.arange(len(list_of_user_activities)))) activities_labels = dict(zip(np.arange(len(list_of_user_activities)), list_of_user_activities)) activities_world = dict(zip(list_of_worlds, np.arange(len(list_of_worlds)))) assess_titles = list(set(train[train['type'] == 'Assessment']['title'].value_counts().index).union( set(test[test['type'] == 'Assessment']['title'].value_counts().index))) # replace the text titles with numbers train['title'] = train['title'].map(activities_map) test['title'] = test['title'].map(activities_map) train['world'] = train['world'].map(activities_world) test['world'] = test['world'].map(activities_world) train_labels['title'] = train_labels['title'].map(activities_map) win_code = dict(zip(activities_map.values(), (4100 * np.ones(len(activities_map))).astype('int'))) win_code[activities_map['Bird Measurer (Assessment)']] = 4110 # convert text into datetime train['timestamp'] = pd.to_datetime(train['timestamp']) test['timestamp'] = pd.to_datetime(test['timestamp']) return train, test, train_labels, win_code, list_of_user_activities, list_of_event_code, activities_labels, assess_titles, list_of_event_id, all_title_event_code # this is the function that convert the raw data into processed features def get_data(user_sample, test_set=False): # Constants and parameters declaration last_activity = 0 user_activities_count = {'Clip': 0, 'Activity': 0, 'Assessment': 0, 'Game': 0} # new features: time spent in each activity last_session_time_sec = 0 accuracy_groups = {0: 0, 1: 0, 2: 0, 3: 0} all_assessments = [] accumulated_accuracy_group = 0 accumulated_accuracy = 0 accumulated_correct_attempts = 0 accumulated_uncorrect_attempts = 0 accumulated_actions = 0 counter = 0 time_first_activity = float(user_sample['timestamp'].values[0]) durations = [] last_accuracy_title = {'acc_' + title: -1 for title in assess_titles} event_code_count: Dict[str, int] = {ev: 0 for ev in list_of_event_code} event_id_count: Dict[str, int] = {eve: 0 for eve in list_of_event_id} title_count: Dict[str, int] = {eve: 0 for eve in activities_labels.values()} title_event_code_count: Dict[str, int] = {t_eve: 0 for t_eve in all_title_event_code} # last features sessions_count = 0 # itarates through each session of one instalation_id for i, session in user_sample.groupby('game_session', sort=False): # i = game_session_id # session is a DataFrame that contain only one game_session # get some sessions information session_type = session['type'].iloc[0] session_title = session['title'].iloc[0] session_title_text = activities_labels[session_title] # for each assessment, and only this kind off session, the features below are processed # and a register are generated if (session_type == 'Assessment') & (test_set or len(session) > 1): # search for event_code 4100, that represents the assessments trial all_attempts = session.query(f'event_code == {win_code[session_title]}') # then, check the numbers of wins and the number of losses true_attempts = all_attempts['event_data'].str.contains('true').sum() false_attempts = all_attempts['event_data'].str.contains('false').sum() # copy a dict to use as feature template, it's initialized with some itens: # {'Clip':0, 'Activity': 0, 'Assessment': 0, 'Game':0} features = user_activities_count.copy() features.update(last_accuracy_title.copy()) features.update(event_code_count.copy()) features.update(event_id_count.copy()) features.update(title_count.copy()) features.update(title_event_code_count.copy()) features.update(last_accuracy_title.copy()) features['installation_session_count'] = sessions_count variety_features = [('var_event_code', event_code_count), ('var_event_id', event_id_count), ('var_title', title_count), ('var_title_event_code', title_event_code_count)] for name, dict_counts in variety_features: arr = np.array(list(dict_counts.values())) features[name] = np.count_nonzero(arr) # get installation_id for aggregated features features['installation_id'] = session['installation_id'].iloc[-1] # add title as feature, remembering that title represents the name of the game features['session_title'] = session['title'].iloc[0] # the 4 lines below add the feature of the history of the trials of this player # this is based on the all time attempts so far, at the moment of this assessment features['accumulated_correct_attempts'] = accumulated_correct_attempts features['accumulated_uncorrect_attempts'] = accumulated_uncorrect_attempts accumulated_correct_attempts += true_attempts accumulated_uncorrect_attempts += false_attempts # the time spent in the app so far if durations == []: features['duration_mean'] = 0 features['duration_std'] = 0 else: features['duration_mean'] = np.mean(durations) features['duration_std'] = np.std(durations) durations.append((session.iloc[-1, 2] - session.iloc[0, 2]).seconds) # the accurace is the all time wins divided by the all time attempts features['accumulated_accuracy'] = accumulated_accuracy / counter if counter > 0 else 0 accuracy = true_attempts / (true_attempts + false_attempts) if (true_attempts + false_attempts) != 0 else 0 accumulated_accuracy += accuracy last_accuracy_title['acc_' + session_title_text] = accuracy # a feature of the current accuracy categorized # it is a counter of how many times this player was in each accuracy group if accuracy == 0: features['accuracy_group'] = 0 elif accuracy == 1: features['accuracy_group'] = 3 elif accuracy == 0.5: features['accuracy_group'] = 2 else: features['accuracy_group'] = 1 features.update(accuracy_groups) accuracy_groups[features['accuracy_group']] += 1 # mean of the all accuracy groups of this player features['accumulated_accuracy_group'] = accumulated_accuracy_group / counter if counter > 0 else 0 accumulated_accuracy_group += features['accuracy_group'] # how many actions the player has done so far, it is initialized as 0 and updated some lines below features['accumulated_actions'] = accumulated_actions # there are some conditions to allow this features to be inserted in the datasets # if it's a test set, all sessions belong to the final dataset # it it's a train, needs to be passed throught this clausule: session.query(f'event_code == {win_code[session_title]}') # that means, must exist an event_code 4100 or 4110 if test_set: all_assessments.append(features) elif true_attempts + false_attempts > 0: all_assessments.append(features) counter += 1 sessions_count += 1 # this piece counts how many actions was made in each event_code so far def update_counters(counter: dict, col: str): num_of_session_count = Counter(session[col]) for k in num_of_session_count.keys(): x = k if col == 'title': x = activities_labels[k] counter[x] += num_of_session_count[k] return counter event_code_count = update_counters(event_code_count, "event_code") event_id_count = update_counters(event_id_count, "event_id") title_count = update_counters(title_count, 'title') title_event_code_count = update_counters(title_event_code_count, 'title_event_code') # counts how many actions the player has done so far, used in the feature of the same name accumulated_actions += len(session) if last_activity != session_type: user_activities_count[session_type] += 1 last_activitiy = session_type # if it't the test_set, only the last assessment must be predicted, the previous are scraped if test_set: return all_assessments[-1] # in the train_set, all assessments goes to the dataset return all_assessments # data split def get_train_and_test(train_test): compiled_train = [] compiled_test = [] for i, (ins_id, user_sample) in tqdm(enumerate(train.groupby('installation_id', sort = False)), total = 17000): compiled_train += get_data(user_sample) for ins_id, user_sample in tqdm(test.groupby('installation_id', sort=False), total=1000): test_data = get_data(user_sample, test_set=True) compiled_test.append(test_data) reduce_train = pd.DataFrame(compiled_train) reduce_test = pd.DataFrame(compiled_test) categoricals = ['session_title'] return reduce_train, reduce_test, categoricals class Base_Model(object): def __init__(self, train_df, test_df, features, categoricals = [], n_splits = 5, verbose = True): self.train_df = train_df self.test_df = test_df self.features = features self.categoricals = categoricals self.target = 'accuracy_group' self.cv = self.get_cv() self.verbose = verbose self.params = self.get_params() self.y_pred, self.score, self.model = self.fit() def train_model(self, train_set, val_set): raise NotImplementedError def get_cv(self): cv = StratifiedKFold(n_splits = slef.n_splits, shuffle = True, random_state = 42) return cv.split(self.train_df, self.train_df[self.target]) def get_params(self): raise NotImplementedError def convert_dataset(self, x_train, y_train, x_val, y_val): raise NotImplementedError def convert_x(self, x): return x def fit(self): oof_pred = np.zeros((len(reduce_train), )) y_pred = np.zeros((len(reduce_test), )) for fold, (train_idx, val_idx) in enumerate(self.cv): x_train, x_val = self.train_df[self.features].iloc[train_idx], self.train_df[self.features].iloc[val_idx] y_train, y_val = self.train_df[self.target][train_idx], self.train_df[self.target][val_idx] train_set, val_set = self.convert_dataset(x_train, y_train, x_val, y_val) model = self.train_model(train_set, val_set) conv_x_val = self.convert_x(x_val) oof_pred[val_idx] = model.predict(conv_x_val).reshape(oof_pred[val_idx].shape) x_test = self.convert_x(self.test_df[self.features]) y_pred += model.predict(x_test).reshape(y_pred.shape) / self.n_splits print('Partial score of fold {} is: {}'.format(fold, eval_qwk_lgb_regr(y_val, oof_pred[val_idx])[1])) _, loss_score, _ = eval_qwk_lgb_regr(self.train_df[self.target], oof_pred) if self.verbose: print('Our oof cohen kappa score is: ', loss_score) return y_pred, loss_score, model # light GBM model class Lgb_Model(Base_Model): def train_model(self, train_set, val_set): verbosity = 100 if self.verbose else 0 return lgb.train(self.params, train_set, valid_sets=[train_set, val_set], verbose_eval=verbosity) def convert_dataset(self, x_train, y_train, x_val, y_val): train_set = lgb.Dataset(x_train, y_train, categorical_feature=self.categoricals) val_set = lgb.Dataset(x_val, y_val, categorical_feature=self.categoricals) return train_set, val_set def get_params(self): params = {'n_estimators': 5000, 'boosting_type': 'gbdt', 'objective': 'regression', 'metric': 'rmse', 'subsample': 0.75, 'subsample_freq': 1, 'learning_rate': 0.01, 'feature_fraction': 0.9, 'max_depth': 15, 'lambda_l1': 1, 'lambda_l2': 1, 'early_stopping_rounds': 100 } return params class Xgb_Model(Base_Model): def train_model(self, train_set, val_set): verbosity = 100 if self.verbose else 0 return xgb.train(self.params, train_set, num_boost_round = 5000, evals = [(train_set, 'train'), (val_set, 'val')], verbose_eval = verbosity, early_stopping_rounds = 100) def convert_dataset(self, x_train, y_train, x_val, y_val): train_set = xgb.DMatrix(x_train, y_val) val_set = xgb.DMatrix(x_val, y_val) return train_set, val_set def convert_x(self, x): return xgb.DMatrix(x) def get_params(self): params = {'colsample_bytree': 0.8, 'learning_rate': 0.01, 'max_depth': 10, 'subsample': 1, 'objective': 'reg:squarederror', # 'eval_metric':'rmse', 'min_child_weight': 3, 'gamma': 0.25, 'n_estimators': 5000} return params # CatBoost Model class Catb_Model(Base_Model): def train_model(self, train_set, val_set): verbosity = 100 if self.verbose else 0 clf = CatBoostRegressor(**self.params) clf.fit(train_set['X'], train_set['y'], eval_set = (val_set['X'], val_set['y']), verbose = verbosity, cat_features = self.categoricals) return clf def convert_dataset(self, x_train, y_train, x_val, y_val): train_set = {'X' : x_train, 'y' : y_val} val_set = {'X': x_val, 'y' : y_val} return train_set, val_set def get_params(self): params = {'loss_function': 'RMSE', 'task_type': "CPU", 'iterations': 5000, 'od_type': "Iter", 'depth': 10, 'colsample_bylevel': 0.5, 'early_stopping_rounds': 300, 'l2_leaf_reg': 18, 'random_seed': 42, 'use_best_model': True } return params class Nn_Model(Base_Model): def __init__(self, train_df, test_df, features, categoricals = [], n_split = 5, verbose = True): features = features.copy() if len(categoricals) > 0 : for cat in categoricals: enc = OneHotEncoder() train_cats = enc.fit_transform(train_df[[cat]]) test_cats = enc.transform(test_df[[cat]]) cal_cols = ['{}_{}'.format(cat, str(col)) for col in enc.activate_features_] features += cat_cols train_cats = pd.DataFrame(train_cats.toarray(), columns = cat_cols) test_cats = pd.DataFrame(test_cats.toarray(), columns = cat_cols) train_df = pd.concat([train_df, train_cats], axis = 1) test_df = pd.concat([test_df, test_cats], axis = 1) scaler = MinMaxScaler() train_df[features] = scaler.fit_transform(train_df[features]) test_df[features] = scaler.fit_transform(test_df[features]) print(train_df[features].shape) super().__init__(train_df, test_df, features, categoricals, n_splits, verbose) def train_model(self, train_set, val_set): verbosity = 100 if self.verbose else 0 model = tf.keras.models.Sequential([ tf.keras.layers.Input(shape = (train_set['X'].shape[1],)) tf.keras.layers.Dense(200, activation = 'relu') tf.keras.layers.LayerNormalization(), tf.keras.layers.Dropout(0.3), tf.keras.layers.Dense(100, activation = 'relu'), tf.keras.layers.LayerNormalization(), tf.keras.layers.Dropout(0.3), tf.keras.layers.Dense(50, activation = 'relu'), tf.keras.layers.LayerNormalization(), tf.keras.layers.Dropout(0.3), tf.keras.layers.Dense(25, activation = 'relu'), tf.keras.layers.LayerNormalization(), tf.keras.layers.Dropout(0.3), tf.keras.layers.Dense(1, activation = 'relu') ]) model.compile(optimizer = tf.keras.optimizers.Adam(learnig_rate = 4e-4), loss = 'mse') print(model.summary()) save_best = tf.keras.callbacks.ModelCheckpoint('nn_model.w8', save_weights_only=True, save_best_only=True, verbose=1) early_stop = tf.keras.callbacks.EarlyStopping(patience=20) model.fit(train_set['X'], train_set['y'], validation_data=(val_set['X'], val_set['y']), epochs=100, callbacks=[save_best, early_stop]) model.load_weights('nn_model.w8') return model def convert_dataset(self, x_train, y_train, x_val, y_val): train_set = {'X': x_train, 'y': y_train} val_set = {'X': x_val, 'y': y_val} return train_set, val_set def get_params(self): return None class Cnn_Model(Base_Model): def __init__(self, train_df, test_df, features, categoricals = [], n_splits = 5, verbose = True): features = features.copy() if len(categoricals) > 0: for cat in categoricals: enc = OneHotEncoder() train_cats = enc.fit_transform(train_df[[cat]]) test_cats = enc.transform(test_df[[cat]]) cat_cols = ['{}_{}'.format(cat, str(col)) for col in enc.active_features_] features += cat_cols train_cats = pd.DataFrame(train_cats.toarray(), columns=cat_cols) test_cats = pd.DataFrame(test_cats.toarray(), columns=cat_cols) train_df = pd.concat([train_df, train_cats], axis=1) test_df = pd.concat([test_df, test_cats], axis=1) # scaler scalar = MinMaxScaler() train_df[features] = scalar.fit_transform(train_df[features]) test_df[features] = scalar.transform(test_df[features]) self.create_feat_2d(features) super().__init__(train_df, test_df, features, categoricals, n_splits, verbose) def create_feat_2d(self, features, n_feats_repeat=50): self.n_feats = len(features) self.n_feats_repeat = n_feats_repeat self.mask = np.zeros((self.n_feats_repeat, self.n_feats), dtype=np.int32) for i in range(self.n_feats_repeat): l = list(range(self.n_feats)) for j in range(self.n_feats): c = l.pop(choice(range(len(l)))) self.mask[i, j] = c self.mask = tf.convert_to_tensor(self.mask) print(self.mask.shape) def train_model(self, train_set, val_set): verbosity = 100 if self.verbose else 0 inp = tf.keras.layers.Input(shape=(self.n_feats)) x = tf.keras.layers.Lambda(lambda x: tf.gather(x, self.mask, axis=1))(inp) x = tf.keras.layers.Reshape((self.n_feats_repeat, self.n_feats, 1))(x) x = tf.keras.layers.Conv2D(18, (50, 50), strides=50, activation='relu')(x) x = tf.keras.layers.Flatten()(x) # x = tf.keras.layers.Dense(200, activation='relu')(x) # x = tf.keras.layers.LayerNormalization()(x) # x = tf.keras.layers.Dropout(0.3)(x) x = tf.keras.layers.Dense(100, activation='relu')(x) x = tf.keras.layers.LayerNormalization()(x) x = tf.keras.layers.Dropout(0.3)(x) x = tf.keras.layers.Dense(50, activation='relu')(x) x = tf.keras.layers.LayerNormalization()(x) x = tf.keras.layers.Dropout(0.3)(x) out = tf.keras.layers.Dense(1)(x) model = tf.keras.Model(inp, out) model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4), loss='mse') print(model.summary()) save_best = tf.keras.callbacks.ModelCheckpoint('nn_model.w8', save_weights_only=True, save_best_only=True, verbose=1) early_stop = tf.keras.callbacks.EarlyStopping(patience=20) model.fit(train_set['X'], train_set['y'], validation_data=(val_set['X'], val_set['y']), epochs=100, callbacks=[save_best, early_stop]) model.load_weights('nn_model.w8') return model def convert_dataset(self, x_train, y_train, x_val, y_val): train_set = {'X': x_train, 'y': y_train} val_set = {'X': x_val, 'y': y_val} return train_set, val_set def get_params(self): return None # read data train, test, train_labels, specs, sample_submission = read_data() # get usefull dict with maping encode train, test, train_labels, win_code, list_of_user_activities, list_of_event_code, activities_labels, assess_titles, list_of_event_id, all_title_event_code = encode_title( train, test, train_labels) # tranform function to get the train and test set reduce_train, reduce_test, categoricals = get_train_and_test(train, test) # feature engineering function features = reduce_train.loc[(reduce_train.sum(axis=1) != 0), (reduce_train.sum(axis=0) != 0)].columns # delete useless columns features = [x for x in features if x not in ['accuracy_group', 'installation_id']] counter = 0 to_remove = [] for feat_a in features: for feat_b in features: if feat_a != feat_b and feat_a not in to_remove and feat_b not in to_remove: c = np.corrcoef(reduce_train[feat_a], reduce_train[feat_b])[0][1] if c > 0.995: counter += 1 to_remove.append(feat_b) print('{}: FEAT_A: {} FEAT_B: {} - Correlation: {}'.format(counter, feat_a, feat_b, c)) to_exclude = [] ajusted_test = reduce_test.copy() for feature in ajusted_test.columns: if feature not in ['accuracy_group', 'installation_id', 'accuracy_group', 'session_title']: data = reduce_train[feature] train_mean = data.mean() data = ajusted_test[feature] test_mean = data.mean() try: error = stract_hists(feature, adjust=True) ajust_factor = train_mean / test_mean if ajust_factor > 10 or ajust_factor < 0.1:# or error > 0.01: to_exclude.append(feature) print(feature, train_mean, test_mean, error) else: ajusted_test[feature] *= ajust_factor except: to_exclude.append(feature) print(feature, train_mean, test_mean) features = [x for x in features if x not in (to_exclude + to_remove)] # modeling #cat_model = Catb_Model(reduce_train, ajusted_test, features, categoricals=categoricals) lgb_model = Lgb_Model(reduce_train, ajusted_test, features, categoricals=categoricals) xgb_model = Xgb_Model(reduce_train, ajusted_test, features, categoricals=categoricals) #cnn_model = Cnn_Model(reduce_train, ajusted_test, features, categoricals=categoricals) nn_model = Nn_Model(reduce_train, ajusted_test, features, categoricals=categoricals) weights = {'lbg': 0.60, 'cat': 0, 'xgb': 0.20, 'nn': 0.20} final_pred = (lgb_model.y_pred * weights['lbg']) + (xgb_model.y_pred * weights['xgb']) + (nn_model.y_pred * weights['nn']) #final_pred = cnn_model.y_pred print(final_pred.shape) dist = Counter(reduce_train['accuracy_group']) for k in dist: dist[k] /= len(reduce_train) reduce_train['accuracy_group'].hist() acum = 0 bound = {} for i in range(3): acum += dist[i] bound[i] = np.percentile(final_pred, acum * 100) print(bound) def classify(x): if x <= bound[0]: return 0 elif x <= bound[1]: return 1 elif x <= bound[2]: return 2 else: return 3 final_pred = np.array(list(map(classify, final_pred))) sample_submission['accuracy_group'] = final_pred.astype(int) sample_submission.to_csv('submission.csv', index=False) sample_submission['accuracy_group'].value_counts(normalize=True) sample_submission.to_csv('submission_1', header = False, index = False)
11492324
import pytest from collections import namedtuple from stack.kvm import Hypervisor, VmException from unittest.mock import create_autospec, patch from stack.commands import DatabaseConnection from stack.argument_processors.vm import VmArgProcessor from stack.commands.set.host.power.imp_kvm import Implementation from stack.commands.set.host.power import Command class TestKVMImp: mock_tuple = namedtuple('output', 'out debug success') def mock_vm_exception(self, *args): raise VmException('Oh no something went wrong!') @pytest.fixture def kvm_imp(self): """ A fixture for mocking the Implementation instance """ mock_command = create_autospec( spec = Command, instance = True ) mock_command.db = create_autospec( spec = DatabaseConnection, spec_set = True, instance = True ) return Implementation(mock_command) @patch.object(target = Implementation, attribute = 'get_hypervisor_by_name', autospec = True) def test_no_kvm_host(self, mock_get_hypervisor_name, kvm_imp): mock_get_hypervisor_name.return_value = None output = kvm_imp.run(args = ['foo', 'on', self.mock_tuple]) assert output == self.mock_tuple('', 'No KVM host specified for virtual machine "foo"', False) @patch('stack.kvm.Hypervisor', autospec = True) def test_kvm_imp_on(self, mock_hypervisor, kvm_imp): hypervisor = mock_hypervisor.return_value.__enter__.return_value hypervisor.start_domain.return_value = None output = kvm_imp.run(args = ['foo', 'on', self.mock_tuple]) hypervisor.start_domain.assert_called_once_with('foo') assert output == self.mock_tuple('', '', True) @patch('stack.kvm.Hypervisor', autospec = True) def test_kvm_imp_off(self, mock_hypervisor, kvm_imp): hypervisor = mock_hypervisor.return_value.__enter__.return_value hypervisor.stop_domain.return_value = None output = kvm_imp.run(args = ['foo', 'off', self.mock_tuple]) hypervisor.stop_domain.assert_called_once_with('foo') assert output == self.mock_tuple('', '', True) @patch('stack.kvm.Hypervisor', autospec = True) def test_kvm_imp_reset(self, mock_hypervisor, kvm_imp): hypervisor = mock_hypervisor.return_value.__enter__.return_value hypervisor.start_domain.return_value = None hypervisor.stop_domain.return_value = None output = kvm_imp.run(args = ['foo', 'reset', self.mock_tuple]) hypervisor.stop_domain.assert_called_once_with('foo') hypervisor.start_domain.assert_called_once_with('foo') assert output == self.mock_tuple('', '', True) VM_STATUS_DEFINED = [ ([{'status': 'on'}], mock_tuple('Chassis Power is on', '', True)), ([{'status': 'off'}], mock_tuple('Chassis Power is off', '', True)), ([{'status': 'undefined'}], mock_tuple('', f'Cannot find host foo defined on hypervisor', False)), ([{'status': 'Connection failed'}], mock_tuple('', f'Cannot find host foo defined on hypervisor', False)) ] @patch('stack.kvm.Hypervisor', autospec = True) @pytest.mark.parametrize('return_status, output', VM_STATUS_DEFINED) def test_kvm_imp_status(self, mock_hypervisor, kvm_imp, return_status, output): kvm_imp.owner.call.return_value = return_status run_output = kvm_imp.run(args = ['foo', 'status', self.mock_tuple]) kvm_imp.owner.call.assert_called_once() assert all([ output.out in run_output.out, output.debug in run_output.debug, output.success == run_output.success ]) @patch('stack.kvm.Hypervisor', autospec = True) def test_kvm_imp_vm_exception(self, mock_hypervisor, kvm_imp): hypervisor = mock_hypervisor.return_value.__enter__.return_value hypervisor.start_domain.side_effect = self.mock_vm_exception output = kvm_imp.run(args = ['foo', 'on', self.mock_tuple]) hypervisor.start_domain.assert_called_once_with('foo') assert output == self.mock_tuple('', 'Oh no something went wrong!', False)
11492346
import argparse import collections import os import pickle import pytest import re import signal import subprocess import sys import time import numpy as np import cntk as C TIMEOUT_SECONDS = 300 NUM_WORKERS = 4 NUM_BATCHES = 10 BATCH_SIZE_PER_WORKER = 20 def mpiexec_execute(script, mpiexec_params, params, timeout_seconds=TIMEOUT_SECONDS): cmd = ['mpiexec'] + mpiexec_params + ['python', script] + params p = subprocess.Popen(cmd, stdout=subprocess.PIPE) if sys.version_info[0] < 3: out = p.communicate()[0] else: try: out = p.communicate(timeout=timeout_seconds)[0] # in case we have a hang except subprocess.TimeoutExpired: os.kill(p.pid, signal.CTRL_C_EVENT) raise RuntimeError('Timeout in mpiexec, possibly hang') str_out = out.decode(sys.getdefaultencoding()) return str_out BlockMomentumConfig = collections.namedtuple('BlockMomentumConfig', 'block_momentum_as_time_constant block_learning_rate block_size distributed_after') DataParallelConfig = collections.namedtuple('DataParallelConfig', 'num_quantization_bits distributed_after') class SimpleTrainer: def __init__(self, mode, config): self.create_model() self.create_trainer(mode, config) def create_model(self): self.input_dim = 1000 self.embed_dim = 30 i = C.input_variable((self.input_dim,), is_sparse=True) self.p = C.parameter(shape=(self.input_dim, self.embed_dim), init=1) o = C.times(i, self.p) self.z = C.reduce_sum(o) def create_trainer(self, mode, config): learner = self.create_distributed_learner(mode, config) self.trainer = C.Trainer(self.z, (self.z, None), learner, []) if learner else None def create_distributed_learner(self, mode, config): local_learner = C.sgd(self.z.parameters, C.learning_parameter_schedule_per_sample(0.01)) try: if mode == 'data_parallel': if config is None: config = DataParallelConfig(num_quantization_bits=32, distributed_after=0) learner = C.data_parallel_distributed_learner(local_learner, num_quantization_bits=config.num_quantization_bits, distributed_after=config.distributed_after) elif mode == 'block_momentum': if config is None: # the default config to match data parallel SGD config = BlockMomentumConfig(block_momentum_as_time_constant=0, block_learning_rate=1, block_size=NUM_WORKERS, distributed_after=0) learner = C.block_momentum_distributed_learner(local_learner, block_momentum_as_time_constant=config.block_momentum_as_time_constant, block_learning_rate=config.block_learning_rate, block_size=config.block_size, distributed_after=config.distributed_after) else: learner = local_learner except RuntimeError: learner = None return learner def train_minibatch(self, input_indices): data = C.Value.one_hot(input_indices, num_classes=self.input_dim) self.trainer.train_minibatch(data) def set_np_random_seed(rank, batch): np.random.seed(rank + 10 * batch) def distributed_worker(outdir, gpu, mode, config): if gpu: # test with only one GPU C.try_set_default_device(C.gpu(0)) else: # CPU sparse aggregation is not implemented, so turn it off # note we only need to explicitly do this when running with CPU device on a GPU build # For CPU build it's disabled by default C.cntk_py.use_sparse_gradient_aggregation_in_data_parallel_sgd(False) trainer = SimpleTrainer(mode, config) for batch in range(NUM_BATCHES): set_np_random_seed(C.Communicator.rank(), batch) indices = (np.random.random((BATCH_SIZE_PER_WORKER,))*(trainer.input_dim-1)).astype(np.int) trainer.train_minibatch(indices) checkpoint_file = os.path.join(outdir, mode+str(batch)) trainer.trainer.save_checkpoint(checkpoint_file) trainer.trainer.restore_from_checkpoint(checkpoint_file) # save a checkpoint to force sync after last minibatch trainer.trainer.save_checkpoint(os.path.join(outdir, mode+'_last')) np.save(os.path.join(outdir, mode+str(C.Communicator.rank())), trainer.p.value) TRAINING_SETTINGS = [ ('data_parallel', None), ('block_momentum', None), ('block_momentum', BlockMomentumConfig(block_momentum_as_time_constant=4000, block_learning_rate=2, block_size=NUM_WORKERS*BATCH_SIZE_PER_WORKER*3, distributed_after=NUM_WORKERS*BATCH_SIZE_PER_WORKER*2)), ('data_parallel', DataParallelConfig(num_quantization_bits=1, distributed_after=NUM_WORKERS*BATCH_SIZE_PER_WORKER*2)), ] @pytest.mark.parametrize("mode, config", TRAINING_SETTINGS) def test_distributed_training_accuracy(tmpdir, device_id, mode, config): ref_trainer = SimpleTrainer(None, None) # test if mode is available if not ref_trainer.create_distributed_learner(mode, config): pytest.skip("unsupported distributed learner mode") # run distributed training and check if all workers get the same model launch_args = ['--outputdir', str(tmpdir), '--mode', mode] if config: config_filename = os.path.join(str(tmpdir),'config.pkl') with open(config_filename, 'wb') as pkl: pickle.dump(config, pkl) launch_args += ['--config', config_filename] if device_id >= 0: launch_args += ['--gpu'] mpiexec_execute(__file__, ['-n', str(NUM_WORKERS)], launch_args) p0 = np.load(os.path.join(str(tmpdir), mode+'0.npy')) for rank in range(NUM_WORKERS): p = np.load(os.path.join(str(tmpdir), mode+str(rank)+'.npy')) assert np.allclose(p0, p) # only compares with single worker with default config if config is not None: return # reference training on single worker, by concatenating data on all workers for batch in range(NUM_BATCHES): indices = None for rank in range(NUM_WORKERS): set_np_random_seed(rank, batch) rank_indices = (np.random.random((BATCH_SIZE_PER_WORKER,))*(ref_trainer.input_dim-1)).astype(np.int) indices = np.concatenate([indices, rank_indices]) if indices is not None else rank_indices ref_trainer.train_minibatch(indices) assert np.allclose(p0, ref_trainer.p.value) #mpiexec entrance if __name__=='__main__': parser = argparse.ArgumentParser() parser.add_argument('-outputdir', '--outputdir') parser.add_argument('-mode', '--mode') parser.add_argument('-gpu', '--gpu', action='store_true') parser.add_argument('-config', '--config', required=False, default=None) args = vars(parser.parse_args()) config = None if args['config'] is not None: with open(args['config'], 'rb') as pkl: config = pickle.load(pkl) distributed_worker(args['outputdir'], args['gpu'], args['mode'], config) C.Communicator.finalize()
11492350
name = 'medcat' # Hacky patch to the built-in copy module coz otherwise, thinc.config.Config.copy will fail on Python <= 3.6. # (fixed in python 3.7 https://docs.python.org/3/whatsnew/3.7.html#re) import sys # noqa if sys.version_info.major == 3 and sys.version_info.minor <= 6: import copy import re copy._deepcopy_dispatch[type(re.compile(''))] = lambda r, _: r # type: ignore
11492416
from django.urls import include from django.views import defaults __all__ = ['handler400', 'handler403', 'handler404', 'handler500', 'include'] handler400 = defaults.bad_request handler403 = defaults.permission_denied handler404 = defaults.page_not_found handler500 = defaults.server_error
11492429
import tensorflow as tf import numpy as np import coremltools as ct from coremltools.models.neural_network import quantization_utils import coremltools.proto.FeatureTypes_pb2 as FeatureTypes_pb2 path = "/Users/william/Downloads/AttGAN_384/generator.pb" # Load the protobuf file from the disk and parse it to retrieve the # graph_def with tf.io.gfile.GFile(path, "rb") as f: graph_def = tf.compat.v1.GraphDef() graph_def.ParseFromString(f.read()) # Import the graph_def into a new Graph with tf.Graph().as_default() as graph: tf.import_graph_def(graph_def, name="generator") x = np.random.rand(1, 384, 384, 3) y = np.ones([1, 13], dtype=float) with tf.compat.v1.Session(graph = graph) as sess: tf_out = sess.run('generator/xb:0', feed_dict={'generator/xa:0': x, 'generator/b_:0':y}) mlmodel = ct.convert(graph, source="tensorflow", inputs=[ct.TensorType(shape=x.shape), ct.TensorType(shape=y.shape)]) model_fp8 = quantization_utils.quantize_weights(mlmodel, nbits=8) spec = model_fp8.get_spec() input = spec.description.input[1] input.type.imageType.colorSpace = FeatureTypes_pb2.ImageFeatureType.RGB input.type.imageType.height = 384 input.type.imageType.width = 384 ct.utils.rename_feature(spec, 'generator/xa', 'input') ct.utils.rename_feature(spec, 'generator/b_', 'style') ct.utils.rename_feature(spec, 'generator/xb', 'output') # ct.utils.save_spec(spec, "attgen8.mlmodel") model_fp8._spec = spec model_fp8.save("attgen8.mlmodel") # Core ML model prediction # coreml_out_dict = model_fp8.predict({"input" : x, "style": y}, useCPUOnly=True) # coreml_out = list(coreml_out_dict.values())[0] # np.testing.assert_allclose(tf_out, coreml_out, rtol=1e-3, atol=1e-2)
11492446
from generativepy.drawing import make_image, setup from generativepy.color import Color from generativepy.geometry import Circle, Square, Text ''' Create bezier curve using the geometry module. ''' def draw(ctx, width, height, frame_no, frame_count): setup(ctx, width, height, width=5, background=Color(0.8)) # Create a circular clip region and draw some squares in it ctx.save() Circle(ctx).of_center_radius((1.9, 1.9), 1).clip() Square(ctx).of_corner_size((1, 1), .8).fill(Color('red')) Square(ctx).of_corner_size((1, 2), .8).fill(Color('green')) Square(ctx).of_corner_size((2, 1), .8).fill(Color('blue')) Square(ctx).of_corner_size((2, 2), .8).fill(Color('black')) ctx.restore() ctx.save() Text(ctx).of("ABC", (1.5, 3.5)).font("Times").size(1.5).align_left().align_top().clip() circles = [(2, 3.8, 'orange'), (2, 4.5, 'cyan'), (3, 3.8, 'green'), (3, 4.5, 'purple'), (4, 3.8, 'yellow'), (4, 4.5, 'blue')] for x, y, color in circles: Circle(ctx).of_center_radius((x, y), 0.7).fill(Color(color)) ctx.restore() make_image("/tmp/geometry-clip.png", draw, 500, 500)
11492484
import cv2 import numpy as np from pyaffx import Affine from _griffig import BoxData, OrthographicImage from ..utility.image import draw_line, get_inference_image from ..utility.model_data import ModelArchitecture class Heatmap: def __init__(self, inference, a_space=None): self.inference = inference if a_space is not None: self.inference.a_space = a_space def calculate_heat(self, reward, size_cropped, size_result): size_reward_center = (reward.shape[1] / 2, reward.shape[2] / 2) scale = self.inference.size_area_cropped[0] / self.inference.size_result[0] * ((size_cropped[0] - 30) / reward.shape[1]) a_space_idx = range(len(self.inference.a_space)) heat_values = np.zeros(size_result[::-1], dtype=np.float) for i in a_space_idx: a = self.inference.a_space[i] rot_mat = cv2.getRotationMatrix2D(size_reward_center, -a * 180.0 / np.pi, scale) rot_mat[0][2] += size_result[0] / 2 - size_reward_center[0] rot_mat[1][2] += size_result[1] / 2 - size_reward_center[1] heat_values += cv2.warpAffine(reward[i], rot_mat, size_result, borderValue=0) norm = (5 * heat_values.max() + len(a_space_idx)) / 6 return (heat_values / norm * 255.0).astype(np.uint8) @staticmethod def get_background(image, use_rgb): if len(image.mat.shape) >= 3 and image.mat.shape[-1] >= 3: if use_rgb: back = cv2.cvtColor(image.mat[:, :, :3], cv2.COLOR_RGB2GRAY) return cv2.cvtColor(back, cv2.COLOR_GRAY2RGB) return cv2.cvtColor(image.mat[:, :, 3:], cv2.COLOR_GRAY2RGB) return cv2.cvtColor(image.mat, cv2.COLOR_GRAY2RGB) def render( self, image: OrthographicImage, object_image: OrthographicImage = None, goal_image: OrthographicImage = None, box_data: BoxData = None, alpha=0.5, use_rgb=True, # Otherwise depth save_path=None, reward_index=None, draw_lateral=False, draw_shifts=False, draw_indices=None, alpha_human=0.0, ): input_images = self.inference.get_input_images(image, box_data) # if goal_image: # input_images += self.inference.get_input_images(goal_image, box_data) if self.inference.model_data.architecture == ModelArchitecture.ActorCritic: estimated_reward, actor_result = self.inference.model(input_images) elif self.inference.model_data.architecture == ModelArchitecture.PlanarSemantic: input_object_images = [get_inference_image(object_image, Affine(a=a), (224, 224), (224, 224), (224, 224), return_mat=True) for a in self.inference.a_space] input_object_images = np.array(input_object_images) / np.iinfo(object_image.mat.dtype).max estimated_grasp_reward, estimated_object_reward = self.inference.model([input_images, [input_object_images]]) estimated_reward = estimated_object_reward * estimated_grasp_reward else: estimated_reward = self.inference.model(input_images) actor_result = None if reward_index is not None: estimated_reward = estimated_reward[reward_index] if estimated_reward.shape[-1] > 4: # self.inference.model_data.output[0] == 'reward+human': estimated_reward = (1 - alpha_human) * estimated_reward[:, :, :, :4] + alpha_human * estimated_reward[:, :, :, 4:] # reward_reduced = np.maximum(estimated_reward, 0) reward_reduced = np.mean(estimated_reward, axis=3) # reward_reduced = estimated_reward[:, :, :, 0] # For heatmapping the actor # reward_reduced = actor_result[:, :, :, 2] # reward_reduced = (reward_reduced - np.min(reward_reduced)) / np.ptp(reward_reduced) # reward_reduced += 0.5 size_cropped = input_images[0].shape[1::-1] size_result = image.mat.shape[1::-1] heat = self.calculate_heat(reward_reduced, size_cropped, size_result) heat = cv2.applyColorMap(heat.astype(np.uint8), cv2.COLORMAP_JET) background = self.get_background(image, use_rgb) if background.dtype == np.uint16: background = background.astype(np.float32) / 255 else: background = background.astype(np.float32) result = (1 - alpha) * background + alpha * heat result = OrthographicImage(result.astype(np.float32), image.pixel_size, image.min_depth, image.max_depth) if draw_indices is not None: self.draw_indices(result, reward_reduced, draw_indices) if draw_lateral: for _ in range(10): index = np.unravel_index(estimated_reward.argmax(), estimated_reward.shape) action = actor_result[index[0], index[1], index[2]] self.draw_lateral(result, estimated_reward.shape, index, action) estimated_reward[np.unravel_index(estimated_reward.argmax(), estimated_reward.shape)] = 0 if draw_shifts: for _ in range(10): self.draw_arrow(result, reward_reduced, np.unravel_index(reward_reduced.argmax(), reward_reduced.shape)) reward_reduced[np.unravel_index(reward_reduced.argmax(), reward_reduced.shape)] = 0 if save_path: cv2.imwrite(str(save_path), result.mat) return result.mat def draw_lateral(self, image: OrthographicImage, reward_shape, index, action): pose = self.inference.pose_from_index(index, reward_shape, image) arrow_color = (255*255, 255*255, 255*255) draw_line(image, pose, Affine(0.0, 0.0), Affine(a=pose.a, b=action[1], c=action[2]) * Affine(0.0, 0.0, -0.14), color=arrow_color, thickness=1) def draw_indices(self, image: OrthographicImage, reward_shape, indices): point_color = (255, 255, 255) for index in indices: pose = self.inference.pose_from_index(index, reward_shape, image) pose.x /= reward_shape[1] / 40 pose.y /= reward_shape[2] / 40 draw_line(image, pose, Affine(-0.001, 0), Affine(0.001, 0), color=point_color, thickness=1) draw_line(image, pose, Affine(0, -0.001), Affine(0, 0.001), color=point_color, thickness=1) def draw_arrow(self, image: OrthographicImage, reward_shape, index): pose = self.inference.pose_from_index(index, reward_shape, image) arrow_color = (255, 255, 255) draw_line(image, pose, Affine(0, 0), Affine(0.036, 0), color=arrow_color, thickness=2) draw_line(image, pose, Affine(0.036, 0.0), Affine(0.026, -0.008), color=arrow_color, thickness=2) draw_line(image, pose, Affine(0.036, 0.0), Affine(0.026, 0.008), color=arrow_color, thickness=2)
11492525
from guided_filter.datasets.google_image import dataFile # -*- coding: utf-8 -*- ## @package guided_filter.results.performance # # Simple performance test. # @author tody # @date 2015/08/26 import os import numpy as np import cv2 import matplotlib.pyplot as plt from guided_filter.io_util.image import loadRGB from guided_filter.cv.image import to32F from guided_filter.util.timer import timing_func, Timer from guided_filter.core.filters import GuidedFilter, FastGuidedFilter from guided_filter.results.results import resultFile ## Bilateral filter class with the same interface with GuidedFilter class. # # Input I for the constructor will be discarded. class BilateralFilter: def __init__(self, I, radius=5, epsilon=0.2): self._epsilon = epsilon self._radius = radius def filter(self, I): return cv2.bilateralFilter(I, 0, self._epsilon, self._radius) ## Peformance test for the input image and the target filter. def performanceTestFilter(C_noise, filter): t = Timer() filter.filter(C_noise) t.stop() return t.seconds() ## Generate filter variations for the target filter class, sigmas. def generateFilterVariations(C_32F, filter_class, sigmas): filters = [] for sigma in sigmas: filters.append(filter_class(C_32F, radius=sigma)) return filters ## Performance test for the target sigmas. def performanceTestSigmas(C_32F, sigmas, filter_types, ax): h, w, cs = C_32F.shape C_noise = np.float32(C_32F + 0.3 * np.random.rand(h, w, cs)) C_noise = np.clip(C_noise, 0.0, 1.0) for type_name, filter_class_color in filter_types.items(): filter_class, color = filter_class_color filters = generateFilterVariations(C_32F, filter_class, sigmas) times = [] for filter in filters: times.append(performanceTestFilter(C_noise, filter)) ax.plot(sigmas, times, label=type_name, color=color) ax.set_xlabel('radius $r$') ax.set_ylabel('time (secs)') ax.legend(bbox_to_anchor=(0.88, 0.8), loc=2) ## Performance test for the image file. def performanceTest(image_file): C_8U = loadRGB(image_file) C_32F = to32F(C_8U) h, w = C_32F.shape[:2] image_size_str = "Image size: %s x %s" %(w, h) fig, axes = plt.subplots(nrows=2, ncols=1, figsize=(8, 8)) fig.subplots_adjust(left=0.1, right=0.7, top=0.86, hspace=0.4) fig.suptitle("Peformance of guided filter\n%s" % image_size_str) filter_types = {"Bilateral Filter": (BilateralFilter, "r"), "Guided Filter": (GuidedFilter, "g"), "Fast Guided Filter": (FastGuidedFilter, "b")} sigmas = range(3, 31, 2) axes[0].set_title('For small radius $r$') performanceTestSigmas(C_32F, sigmas, filter_types, axes[0]) sigmas = range(10, 100, 5) filter_types = {"Guided Filter": (GuidedFilter, "g"), "Fast Guided Filter": (FastGuidedFilter, "b")} axes[1].set_title('For large radius $r$') performanceTestSigmas(C_32F, sigmas, filter_types, axes[1]) result_name = "performance" result_file = resultFile(result_name) plt.savefig(result_file) ## Performance tests for the data names, IDs. def performanceTests(data_names, data_ids): for data_name in data_names: print "Performance tests: %s" % data_name for data_id in data_ids: print "Data ID: %s" % data_id image_file = dataFile(data_name, data_id) performanceTest(image_file) if __name__ == '__main__': data_names = ["flower"] data_ids = [0] performanceTests(data_names, data_ids)
11492528
import collections import csv import glob import os import pandas as pd import cea import cea.config import cea.inputlocator __author__ = "<NAME>" __copyright__ = "Copyright 2018, Architecture and Building Systems - ETH Zurich" __credits__ = ["<NAME>"] __license__ = "MIT" __version__ = "0.1" __maintainer__ = "<NAME>" __email__ = "<EMAIL>" __status__ = "Production" COLUMNS_SCHEDULES = ['DAY', 'HOUR', 'OCCUPANCY', 'APPLIANCES' 'LIGHTING', 'WATER', 'HEATING', 'COOLING', 'PROCESSES', 'SERVERS'] DAY = ['WEEKDAY'] * 24 + ['SATURDAY'] * 24 + ['SUNDAY'] * 24 HOUR = list(range(1, 25)) + list(range(1, 25)) + list(range(1, 25)) def read_cea_schedule(path_to_cea_schedule): """ reader for the files ``locator.get_building_weekly_schedules`` :param str path_to_cea_schedule: path to the cea schedule file to read. (E.g. inputs/building-properties/schedules/B001.csv) :return: schedule data, schedule complementary data """ with open(path_to_cea_schedule) as f: reader = csv.reader(f) for i, row in enumerate(reader): if i == 0: metadata = row[1] elif i == 1: monthly_multiplier = [round(float(x), 2) for x in row[1:]] else: # skip all the other rows break schedule_data = pd.read_csv(path_to_cea_schedule, skiprows=2).T schedule_data = dict(zip(schedule_data.index, schedule_data.values)) schedule_complementary_data = {'METADATA': metadata, 'MONTHLY_MULTIPLIER': monthly_multiplier} return schedule_data, schedule_complementary_data def save_cea_schedule(schedule_data, schedule_complementary_data, path_to_building_schedule): METADATA = ['METADATA'] + [schedule_complementary_data['METADATA']] MULTIPLIER = ['MONTHLY_MULTIPLIER'] + list(schedule_complementary_data['MONTHLY_MULTIPLIER']) COLUMNS_SCHEDULES = schedule_data.keys() RECORDS_SCHEDULES = map(list, zip(*schedule_data.values())) with open(path_to_building_schedule, "w", newline="", encoding="utf-8") as csvfile: csvwriter = csv.writer(csvfile, delimiter=',') csvwriter.writerow(METADATA) csvwriter.writerow(MULTIPLIER) csvwriter.writerow(COLUMNS_SCHEDULES) for row in RECORDS_SCHEDULES: csvwriter.writerow(row) def get_all_schedule_names(schedules_folder): """Get all schedule names from path""" schedule_files = glob.glob(os.path.join(schedules_folder, "*.csv")) return [os.path.splitext(os.path.basename(schedule_file))[0] for schedule_file in schedule_files] # TODO: Replace usages of `read_cea_schedule` to this function def schedule_to_dataframe(schedule_path): out = collections.OrderedDict() with open(schedule_path) as f: reader = csv.reader(f) out['METADATA'] = pd.DataFrame({'metadata': [next(reader)[1]]}) out['MONTHLY_MULTIPLIER'] = pd.DataFrame({m + 1: [round(float(v), 2)] for m, v in enumerate(next(reader)[1:])}, columns=[m for m in range(1, 13)]) # Filter empty columns columns = [col for col in next(reader) if col != ''] schedule_data = pd.read_csv(schedule_path, skiprows=2, usecols=columns).set_index( ['DAY', 'HOUR']).unstack().reindex(['WEEKDAY', 'SATURDAY', 'SUNDAY']) for t, df in schedule_data.groupby(axis=1, level=0, sort=False): df.columns = [i for i in range(1, 25)] out[t] = df.reset_index() return out def schedule_to_file(schedule, schedule_path): schedule_df = pd.DataFrame() metadata = ['METADATA'] multiplier = ['MONTHLY_MULTIPLIER'] for key, data in schedule.items(): if key == 'METADATA': metadata += [schedule['METADATA']['metadata'].iloc[0]] elif key == 'MONTHLY_MULTIPLIER': multiplier += [i for i in schedule['MONTHLY_MULTIPLIER'].iloc[0].values] else: schedule_column_data = data.set_index(['DAY']).reindex(['WEEKDAY', 'SATURDAY', 'SUNDAY']).stack() schedule_column_data.index.names = ['DAY', 'HOUR'] schedule_df[key] = schedule_column_data schedule_df = schedule_df.reset_index() with open(schedule_path, "w", newline="", encoding="utf-8") as csv_file: csv_writer = csv.writer(csv_file, delimiter=',') csv_writer.writerow(metadata) csv_writer.writerow(multiplier) csv_writer.writerow(schedule_df.columns) for row in schedule_df.values: csv_writer.writerow(row) print('Schedule file written to {}'.format(schedule_path)) def main(config): locator = cea.inputlocator.InputLocator(scenario=config.scenario) path_to_building_schedule = locator.get_database_standard_schedules_use('MULTI_RES') # print(read_cea_schedule(path_to_building_schedule)) print(schedule_to_dataframe(path_to_building_schedule)) if __name__ == '__main__': main(cea.config.Configuration())
11492533
import os import pytest from waterbutler.core import streams DUMMY_FILE = os.path.join(os.path.dirname(__file__), 'fixtures/dummy.txt') class TestFileStreamReader: @pytest.mark.asyncio async def test_file_stream_reader(self): with open(DUMMY_FILE, 'r') as fp: reader = streams.FileStreamReader(fp) assert reader.size == 27 data = await reader.read() assert data == 'abcdefghijklmnopqrstuvwxyz\n' at_eof = reader.at_eof() assert not at_eof data = await reader.read() assert data == b'' at_eof = reader.at_eof() assert at_eof reader.close() at_eof = reader.at_eof() assert at_eof with pytest.raises(ValueError): fp.read() @pytest.mark.asyncio async def test_file_stream_reader_after_seek(self): with open(DUMMY_FILE, 'r') as fp: fp.seek(3) reader = streams.FileStreamReader(fp) assert reader.size == 27 # still gives full size assert fp.tell() == 3 # returns to original seek position data = await reader.read() assert data == 'abcdefghijklmnopqrstuvwxyz\n' # always reads full data at_eof = reader.at_eof() assert not at_eof data = await reader.read() assert data == b'' at_eof = reader.at_eof() assert at_eof @pytest.mark.asyncio async def test_file_stream_reader_subset(self): with open(DUMMY_FILE, 'r') as fp: reader = streams.FileStreamReader(fp) data = await reader.read(10) assert data == 'abcdefghij' at_eof = reader.at_eof() assert not at_eof data = await reader.read(2) assert data == 'kl' at_eof = reader.at_eof() assert not at_eof data = await reader.read() assert data == 'mnopqrstuvwxyz\n' at_eof = reader.at_eof() assert not at_eof data = await reader.read() assert data == b'' at_eof = reader.at_eof() assert at_eof class TestPartialFileStreamReader: @pytest.mark.asyncio @pytest.mark.parametrize("byte_range,size,is_partial,content_range,expected", [ ((0, 26), 27, False, 'bytes 0-26/27', 'abcdefghijklmnopqrstuvwxyz\n'), ((0, 5), 6, True, 'bytes 0-5/27', 'abcdef'), ((2, 10), 9, True, 'bytes 2-10/27', 'cdefghijk'), ((20, 26), 7, True, 'bytes 20-26/27', 'uvwxyz\n'), ((2, 2), 1, True, 'bytes 2-2/27', 'c'), ]) async def test_partial_file_stream_reader(self, byte_range, size, is_partial, content_range, expected): with open(DUMMY_FILE, 'r') as fp: reader = streams.PartialFileStreamReader(fp, byte_range) assert reader.size == size assert reader.total_size == 27 assert reader.partial == is_partial assert reader.content_range == content_range data = await reader.read() assert data == expected at_eof = reader.at_eof() assert not at_eof data = await reader.read() assert data == b'' at_eof = reader.at_eof() assert at_eof @pytest.mark.asyncio @pytest.mark.parametrize("byte_range,size,is_partial,content_range,expected", [ ((0, 26), 27, False, 'bytes 0-26/27', 'abcdefghijklmnopqrstuvwxyz\n'), ((0, 5), 6, True, 'bytes 0-5/27', 'abcdef'), ((2, 10), 9, True, 'bytes 2-10/27', 'cdefghijk'), ((20, 26), 7, True, 'bytes 20-26/27', 'uvwxyz\n'), ((2, 2), 1, True, 'bytes 2-2/27', 'c'), ]) async def test_partial_file_stream_reader_with_size(self, byte_range, size, is_partial, content_range, expected): """Test that range is respected even when large size values are passed to ``.read()``.""" with open(DUMMY_FILE, 'r') as fp: reader = streams.PartialFileStreamReader(fp, byte_range) assert reader.size == size assert reader.total_size == 27 assert reader.partial == is_partial assert reader.content_range == content_range data = await reader.read(500) assert data == expected at_eof = reader.at_eof() assert not at_eof data = await reader.read(500) assert data == b'' at_eof = reader.at_eof() assert at_eof
11492549
from colorclass import Color from terminaltables import SingleTable class ArticlePrinter(object): """Class which can be used to print Article stats on the command line""" def __init__(self, article_obj): self.article = article_obj def print_article_stats(self): """This method is called to present overall article stats on a command line.""" table_data = [ [Color('{autocyan}Overall Stats{/autocyan}')], ['Reading time', str(self.article.reading_time) + ' mins'], ['Flesch Reading Ease', self.article.get_flesch_reading_score()], ['Dale Chall Readability Score', self.article.get_dale_chall_reading_score()], ['Paragraphs', self.article.total_paragraphs], ['Avg sentences per paragraph', self.article.avg_sentences_per_para], ['Total sentences in longest paragraph', self.article.len_of_longest_paragraph], ['Sentences', self.article.total_sentences], ['Avg words per sentence', self.article.avg_words_per_sentence], ['Longest sentence', "%s..." % str(self.article.longest_sentence)[0:30]], ['Words in longest sentence', self.article.len_of_longest_sentence], ['Words', self.article.total_words], ['"and" frequency"', self.article.get_and_frequency()], ['Compulsive Hedgers', len(self.article.get_compulsive_hedgers())], ['Intensifiers', len(self.article.get_intensifiers())], ['Vague words', len(self.article.get_vague_words())], ] table_instance = SingleTable(table_data) table_instance.inner_heading_row_border = True table_instance.inner_row_border = True table_instance.justify_columns = {0: 'left', 1: 'center'} print(table_instance.table) self.print_detail() def print_paragraph_stats(self): """This method, along with print_article_stats(), can be called to present paragraph stats on a command line. Ideally first call print_article_stats() and then this method. It shows sentence, average words per sentence, longest sentence, and readability scores (Flesch reading ease and Dale Chall readability scores) of paragraphs. """ sentence_tag = Color('{blue}sentences{/blue}') word_tag = Color('{blue}words{/blue}') avg_word_tag = Color('{blue}Avg words per sentence{/blue}') long_tag = Color('{red}longest{/red}') table_data = [ [Color('{autocyan}Paragraph Stats{/autocyan}')], ['Paragraph #', ''] ] for item, para in enumerate(self.article.paragraphs): sentences = Color('{red}%s{/red}' % str(len(para))) if len(para) > 5 else str(len(para)) avg_words_per_sentence = Color( '{red}%s{/red}' % str(para.avg_words_per_sentence)) if para.avg_words_per_sentence > 25 else str( para.avg_words_per_sentence) table_data.append([item + 1, '{sentences} {sent_tag}. {words} {word_tag}. {avg_words} {avg_word_tag}. ' '"{longest_sent}..." is the {long_tag} sentence.'.format( sentences=sentences, sent_tag=sentence_tag, words=para.total_words, word_tag=word_tag, avg_words=avg_words_per_sentence, avg_word_tag=avg_word_tag, longest_sent=str(para.longest_sentence)[0:10], long_tag=long_tag )]) table_data.append(["", "Flesh Reading score={flesch_reading}, Dale Chall Readability= {dale_chall}".format( flesch_reading=para.get_flesch_reading_score(), dale_chall=para.get_dale_chall_reading_score() )]) table_instance = SingleTable(table_data) table_instance.inner_heading_row_border = True table_instance.inner_row_border = True table_instance.justify_columns = {0: 'center', 1: 'left'} print(table_instance.table) def _print_detail_of(self, words_list, heading, display_count=True): words = [str(word) for word in words_list] if len(words) >= 1: words_unique_list = list(set(words)) format_str = "{word} ({count})" if display_count: msg = '{red} **- %s: %s {/red}\r\n' % (heading, ', '.join(format_str.format(word=str(word), count=words.count(word)) for word in words_unique_list)) else: msg = '{red} **- %s: %s {/red}\r\n' % (heading, ', '.join(word for word in words_unique_list)) print(Color(msg)) def print_detail(self): self._print_detail_of(self.article.get_compulsive_hedgers(), "Compulsive Hedgers", display_count=False) self._print_detail_of(self.article.get_intensifiers(), "Intensifiers", display_count=False) self._print_detail_of(self.article.get_vague_words(), "Vague words") self._print_detail_of(self.article.ten_words_with_most_syllables(), "10 words with most syllables", display_count=False) self._print_detail_of(self.article.get_n_most_repeated_words(20), "20 most repeated words", display_count=False)
11492587
def gcd(u, v): while v > 0: tmp = v v = u % v u = tmp return u t = int(input()) for _ in range(t): a, b, x, y = map(int, input().split()) # gcd definitions says: # gcd (a + m * b, b) = gcd(a, b) # such as m being an integer # hence: # gcd (a + b, b) # gcd (a - b, b) # gcd (a, a + b) # gcd (a, a - b) # must all be equal to gcd(a, b) # thus it means it's possible to # move from (a,b) point to (x, y) point # iff gcd(a, b) == gcd(x, y) if gcd(a, b) == gcd(x, y): print("YES") else: print("NO")
11492593
import torch import numpy as np from torch import nn, optim from .data_helper import get_data_loader history = {} def root_mean_squared_error(y_true, y_pred): """ RMSE implementation. :param y_true: The correct labels :param y_pred: The predicted labels :return: RMSE score """ return np.sqrt(((y_true - y_pred) ** 2).mean()) def data_result_size(data): """ Count all samples in all objects for all workers in a grid search response. :param data: Grid search response data :return: Total number of data samples """ total = 0 for i in range(len(data)): for j in range(len(data[i])): total += data[i][j].shape[0] return total def train(model, data, labels, criterion): """ Train the model with the given data in a federated way. :param model: Model to train :param data: Data to use for training :param labels: Labels to use for loss calculation :param criterion: Criterion to use for loss calculations :return: Loss of this epoch """ model.train() epoch_total = data_result_size(data) running_loss = 0 for i in range(len(data)): # initialize an dedicated optimizer for every worker to prevent errors with adams momentum optimizer = optim.Adam(model.parameters()) for j in range(len(data[i])): # check the location of the data and send the model there worker = data[i][j].location model.send(worker) # train one step optimizer.zero_grad() output = model(data[i][j]) loss = criterion(output, labels[i][j]) loss.backward() optimizer.step() # get the updated model and the loss back from the worker model.get() loss = loss.get() running_loss += loss.item() * data[i][j].shape[0] epoch_loss = running_loss / epoch_total # TODO: add EarlyStopping return epoch_loss def test(model, test_loader, criterion): """ Test the model. :param model: The model to test :param test_loader: DataLoader with the test data :param criterion: Criterion to use for loss calculation :return: Loss of the test """ model.eval() test_loss = 0 with torch.no_grad(): for data, target in test_loader: output = model(data) test_loss += criterion(output, target).item() * data.size(0) test_loss /= len(test_loader.dataset) return test_loss def start_federated_training(model, train_data, train_labels, val_data, val_labels, epochs): """ Start federated training. :param model: Model to train :param train_data: Training data :param train_labels: Training labels :param val_data: Validation data :param val_labels: Validation labels :param epochs: Number of epochs :return: The trained model """ global history torch.manual_seed(1) np.random.seed(51) criterion = nn.L1Loss() # mae # manually track the losses history['epoch'] = [] history['loss'] = [] history['val_loss'] = [] val_loader = get_data_loader(val_data, val_labels) for epoch in range(1, epochs + 1): train_loss = train(model, train_data, train_labels, criterion) val_loss = test(model, val_loader, criterion) print('Epoch: {}/{}\tloss: {:.4f}\tval_loss: {:.4f}'.format(epoch, epochs, train_loss, val_loss)) history['epoch'].append(epoch) history['loss'].append(train_loss) history['val_loss'].append(val_loss) return model def get_model_error(model, test_data, test_labels): """ Calculate the RMSE for the given model and test data. :param model: Model to use :param test_data: test data :param test_labels: test labels :return: RMSE score """ model.eval() with torch.no_grad(): output = model(test_data) rmse = root_mean_squared_error(test_labels, output) return rmse
11492647
import os, os.path, errno, sys, traceback, subprocess import re, html.entities import json import logging import yaml from bs4 import BeautifulSoup from datetime import datetime import requests import urllib.parse import io import gzip import docx import zipfile from urllib.parse import urljoin import inspect import pdfrw from . import admin logging.getLogger("pdfrw").setLevel(logging.CRITICAL) # scraper should be instantiated at class-load time, so that it can rate limit appropriately import scrapelib scraper = scrapelib.Scraper(requests_per_minute=120, retry_attempts=3) scraper.user_agent = "unitedstates/inspectors-general (https://github.com/unitedstates/inspectors-general)" scraper.timeout = 60 class Soft404HttpAdapter(requests.adapters.HTTPAdapter): """Transport adapter that checks all responses against a blacklist of "file not found" pages that are served with 200 status codes.""" SOFT_404_BODY_SIGNATURES = { "si.edu": b"<title>Page Not FoundSmithsonian</title>", } def build_response(self, req, resp): domain = urllib.parse.urlparse(req.url)[1].split(':')[0] base_domain = ".".join(domain.split(".")[-2:]) if base_domain in self.SOFT_404_BODY_SIGNATURES: if resp.getheader("Content-Type") in ["text/html; charset=utf-8", "text/html"]: data = resp.data headers = resp.headers if resp.getheader("Content-Encoding") == "gzip": decompressed_data = gzip.decompress(data) else: decompressed_data = data if resp.getheader("Transfer-Encoding") == "chunked": headers.pop("Transfer-Encoding") body = io.BytesIO(data) resp = requests.packages.urllib3.response.HTTPResponse( body=body, headers=headers, status=resp.status, version=resp.version, reason=resp.reason, strict=resp.strict, preload_content=False, ) if decompressed_data.find(self.SOFT_404_BODY_SIGNATURES[base_domain], 0, 10240) != -1: result = super(Soft404HttpAdapter, self).build_response(req, resp) result.status_code = 404 # tells scrapelib to not retry return result result = super(Soft404HttpAdapter, self).build_response(req, resp) return result scraper.mount("http://www.si.edu/", Soft404HttpAdapter()) scraper.mount("http://si.edu/", Soft404HttpAdapter()) scraper.mount("https://www.si.edu/", Soft404HttpAdapter()) scraper.mount("https://si.edu/", Soft404HttpAdapter()) WHITELIST_INSECURE_DOMAINS = ( "https://www.ncua.gov/", # incomplete chain as of 12/10/2016 ) # Special case handling for governmentattic.org, va.gov, etc.: # These pages are served without an encoding in the HTTP headers, # and with the encoding specified in a <meta> tag inside the document. # See https://gist.github.com/divergentdave/0985ae1f2fd2a7235ccef0d5bbb6aaa3 # for data collection script. META_CHARSETS = { "http://www.usda.gov/oig": "iso-8859-1", "http://www.cftc.gov/": "utf-8", "http://cftc.gov/": "utf-8", "https://www.cftc.gov/": "utf-8", "https://cftc.gov/": "utf-8", "http://www.dodig.mil/": "utf-8", "https://www.dodig.mil/": "utf-8", "https://oig.justice.gov/": "utf-8", "https://www.fca.gov/": "utf-8", "https://fca.gov/": "utf-8", "http://www.fec.gov/": "iso-8859-1", "https://www.fec.gov/": "iso-8859-1", "https://oig.federalreserve.gov/": "iso-8859-1", "http://www.gao.gov/": "utf-8", "http://gao.gov/": "utf-8", "https://www.gao.gov/": "utf-8", "https://gao.gov/": "utf-8", "http://www.governmentattic.org/": "utf-8", "http://governmentattic.org/": "utf-8", "https://www.governmentattic.org/": "utf-8", "https://governmentattic.org/": "utf-8", "https://oig.nasa.gov/": "utf-8", "http://www.nrc.gov/": "utf-8", "https://www.nrc.gov/": "utf-8", "http://oig.pbgc.gov/": "utf-8", "https://oig.pbgc.gov/": "utf-8", "https://www.treasury.gov/tigta/": "iso-8859-1", "http://oig.tva.gov/": "utf-8", "https://oig.tva.gov/": "utf-8", "https://www.va.gov/oig/": "utf-8", } # will pass correct options on to individual scrapers whether # run through ./igs or individually, because argv[1:] is the same def run(run_method, additional=None): cli_options = options() configure_logging(cli_options) if additional: cli_options.update(additional) try: return run_method(cli_options) except Exception as exception: admin.log_exception(exception) # read options from the command line # e.g. ./inspectors/usps.py --since=2012-03-04 --debug # => {"since": "2012-03-04", "debug": True} AVAILABLE_OPTIONS = ( "archive", "bulk", "component", "debug", "dry_run", "end", "ig", "limit", "log", "only", "pages", "quick", "report_id", "safe", "since", "skip_downloaded", "start", "topics", "types", "year", ) def options(): options = {} for arg in sys.argv[1:]: if arg.startswith("--"): if "=" in arg: key, value = arg.split('=') else: key, value = arg, "true" key = key.split("--")[1] key = key.lower() value = value.lower() if key not in AVAILABLE_OPTIONS: print("Unknown option: \"%s\"\n" "The following options are recognized\n" " %s"% (key, ", ".join(AVAILABLE_OPTIONS))) sys.exit(1) if value == 'true': value = True elif value == 'false': value = False options[key] = value return options def configure_logging(options=None): options = {} if not options else options if options.get('debug', False): log_level = "debug" else: log_level = options.get("log", "warn") if log_level not in ["debug", "info", "warn", "error"]: print("Invalid log level (specify: debug, info, warn, error).") sys.exit(1) logging.basicConfig(format='%(message)s', level=log_level.upper()) # download the data at url def download(url, destination=None, options=None, scraper_slug=None): options = {} if not options else options cache = options.get('cache', True) # default to caching binary = options.get('binary', False) # default to assuming text # check cache first if destination and cache and os.path.exists(destination): logging.info("## Cached: (%s, %s)" % (destination, url)) # if a binary file is cached, we're done if binary: return True # otherwise, decode it for return with open(destination, 'r', encoding='utf-8') as f: body = f.read() # otherwise, download from the web else: logging.warn(url) logging.info("## Downloading: %s" % url) if binary: if destination: logging.info("## \tto: %s" % destination) else: raise Exception("A destination path is required for downloading a binary file") try: mkdir_p(os.path.dirname(destination)) verify_options = domain_verify_options(url) scraper.urlretrieve(url, destination, verify=verify_options) except connection_errors() as e: admin.log_http_error(e, url, scraper_slug) return None else: # text try: if destination: logging.info("## \tto: %s" % destination) # Special handler for downloading reports whose server has # misconfigured their HTTPS, and for which no alternative # exists. # This happens very rarely, and scrapelib has a bug with # verification options, so this disables the rate limiting # provided by scrapelib. verify_options = domain_verify_options(url) response = scraper.get(url, verify=verify_options) except connection_errors() as e: admin.log_http_error(e, url, scraper_slug) return None for prefix, charset in META_CHARSETS.items(): if url.startswith(prefix): response.encoding = charset break body = response.text if not isinstance(body, str): raise ValueError("Content not decoded.") # don't allow 0-byte files if (not body) or (not body.strip()): return None # cache content to disk if destination: write(body, destination, binary=binary) # don't return binary content if binary: return True else: # whether from disk or web, unescape HTML entities return unescape(body) def beautifulsoup_from_url(url): caller_filename = inspect.stack()[1][1] caller_scraper = os.path.splitext(os.path.basename(caller_filename))[0] body = download(url, scraper_slug=caller_scraper) if body is None: return None doc = BeautifulSoup(body, "lxml") # Some of the pages will return meta refreshes if doc.find("meta") and doc.find("meta").attrs.get('http-equiv') == 'REFRESH': redirect_url = urljoin(url, doc.find("meta").attrs['content'].split("url=")[1]) return beautifulsoup_from_url(redirect_url) else: return doc def post(url, data=None, headers=None, **kwargs): response = None try: verify_options = domain_verify_options(url) response = scraper.post(url, data=data, headers=headers, verify=verify_options) except connection_errors() as e: admin.log_http_error(e, url) return None return response def resolve_redirect(url): res = scraper.request(method='HEAD', url=url, allow_redirects=False) if "Location" in res.headers: return res.headers["Location"] else: return url def connection_errors(): return (scrapelib.HTTPError, requests.exceptions.ConnectionError, requests.packages.urllib3.exceptions.MaxRetryError) # uses BeautifulSoup to do a naive extraction of text from HTML, # then writes it and returns the /data-relative path. def text_from_html(real_html_path, real_text_path): html = open(real_html_path, encoding='utf-8').read() doc = BeautifulSoup(html, "lxml") for node in doc.findAll(['script', 'style']): node.extract() text = doc.text lines = text.splitlines() for i in range(len(lines)): lines[i] = lines[i].strip() lines = filter(None, lines) text = "\n".join(lines) write(text, real_text_path, binary=False) def domain_verify_options(url): for domain in WHITELIST_INSECURE_DOMAINS: if url.startswith(domain): logging.warn("SKIPPING HTTPS VERIFICATION.") return False return True _tool_present_cache = {} def check_tool_present(*args): if args in _tool_present_cache: return _tool_present_cache[args] try: subprocess.Popen(args, shell=False, stdout=subprocess.DEVNULL, stderr=subprocess.STDOUT).communicate() result = True except FileNotFoundError: result = False _tool_present_cache[args] = result return result # read PDF's directory to determine if we need to decrypt it def check_pdf_decryption(pdf_path): try: doc = pdfrw.PdfReader(pdf_path) return "/Encrypt" in doc except: return False # uses qpdf to decrypt a PDF def decrypt_pdf(source_path, destination_path): if not check_tool_present("qpdf", "--version"): logging.warn("Install qpdf to decrypt PDFs! " "The qpdf executable must be in a directory that is in " "your PATH environment variable.") return False try: subprocess.check_call(["qpdf", "--decrypt", source_path, destination_path], shell=False) return True except subprocess.CalledProcessError as exc: logging.warn("Error decrypting %s:\n\n%s" % (source_path, format_exception(exc))) return False if not os.path.exists(destination_path): logging.warn("PDF not decrypted to %s" % destination_path) return False # uses pdftotext to get text out of PDFs, # then writes it and returns the /data-relative path. def text_from_pdf(real_pdf_path, real_text_path): if not check_tool_present("pdftotext", "-v"): logging.warn("Install pdftotext to extract text! " "The pdftotext executable must be in a directory that is in " "your PATH environment variable.") return try: subprocess.check_call(["pdftotext", "-layout", "-nopgbrk", real_pdf_path, real_text_path], shell=False) except subprocess.CalledProcessError as exc: logging.warn("Error extracting text to %s:\n\n%s" % (real_text_path, format_exception(exc))) return if not os.path.exists(real_text_path): logging.warn("Text not extracted to %s" % real_text_path) def text_from_doc(real_doc_path, real_text_path): if not check_tool_present("abiword", "-?"): logging.warn("Install AbiWord to extract text! " "The abiword executable must be in a directory that is in " "your PATH environment variable.") return try: subprocess.check_call(["abiword", real_doc_path, "--to", "txt"], shell=False) except subprocess.CalledProcessError as exc: logging.warn("Error extracting text to %s:\n\n%s" % (real_text_path, format_exception(exc))) return if not os.path.exists(real_text_path): logging.warn("Text not extracted to %s" % real_text_path) def text_from_docx(real_docx_path, real_text_path): def text_from_paragraphs(paragraphs): return "\n\n".join([paragraph.text for paragraph in paragraphs]) def text_from_tables(tables): return "\n\n".join([text_from_table(table) for table in tables]) def text_from_table(table): return "\n\n".join([text_from_row(row) for row in table.rows]) def text_from_row(row): return "\n\n".join([text_from_doc_or_cell(cell) for cell in row.cells]) def text_from_doc_or_cell(cell): part1 = text_from_paragraphs(cell.paragraphs) part2 = text_from_tables(cell.tables) if part1 and part2: return "%s\n\n%s" % (part1, part2) else: return part1 + part2 try: document = docx.Document(real_docx_path) text = text_from_doc_or_cell(document) write(text, real_text_path, binary=False) except zipfile.BadZipFile as exc: logging.warn("Error extracting text to %s:\n\n%s" % (real_text_path, format_exception(exc))) return None PDF_PAGE_RE = re.compile("Pages: +([0-9]+)\r?\n") PDF_CREATION_DATE_RE = re.compile("CreationDate: +([^\r\n]*)\r?\n") PDF_MOD_DATE_RE = re.compile("ModDate: +([^\r\n]*)\r?\n") PDF_TITLE_RE = re.compile("Title: +([^\r\n]*)\r?\n") PDF_KEYWORDS_RE = re.compile("Keywords: +([^\r\n]*)\r?\n") PDF_AUTHOR_RE = re.compile("Author: +([^\r\n]*)\r?\n") def parse_pdf_datetime(raw): if raw.strip() == "": return None my_datetime = None datetime_formats = [ '%m/%d/%y %H:%M:%S', '%a %b %d %H:%M:%S %Y', '%A, %B %d, %Y %I:%M:%S %p', '%a %b %d %H:%M:%S %Y %Z' ] for datetime_format in datetime_formats: try: my_datetime = datetime.strptime(raw, datetime_format) break except ValueError: pass if my_datetime: return datetime.strftime(my_datetime, '%Y-%m-%d') else: logging.warn('Could not parse PDF date: %s' % raw) return None def metadata_from_pdf(pdf_path): if not check_tool_present("pdfinfo", "-v"): logging.warn("Install pdfinfo to extract metadata! " "The pdfinfo executable must be in a directory that is in " "your PATH environment variable.") return None real_pdf_path = os.path.expandvars(os.path.join(data_dir(), pdf_path)) real_pdf_path = os.path.abspath(real_pdf_path) try: output = subprocess.check_output(["pdfinfo", real_pdf_path], shell=False) output = output.decode('utf-8', errors='replace') except subprocess.CalledProcessError as exc: logging.warn("Error extracting metadata for %s:\n\n%s" % (pdf_path, format_exception(exc))) return None metadata = {} page_match = PDF_PAGE_RE.search(output) if page_match: metadata['page_count'] = int(page_match.group(1)) creation_date_match = PDF_CREATION_DATE_RE.search(output) if creation_date_match: metadata['creation_date'] = parse_pdf_datetime(creation_date_match.group(1)) mod_date_match = PDF_MOD_DATE_RE.search(output) if mod_date_match: metadata['modification_date'] = parse_pdf_datetime(mod_date_match.group(1)) title_match = PDF_TITLE_RE.search(output) if title_match: metadata['title'] = title_match.group(1) keywords_match = PDF_KEYWORDS_RE.search(output) if keywords_match: metadata['keywords'] = keywords_match.group(1) author_match = PDF_AUTHOR_RE.search(output) if author_match: metadata['author'] = author_match.group(1) if metadata: return metadata return None def check_report_url(report_url): try: verify_options = domain_verify_options(report_url) scraper.request(method='HEAD', url=report_url, verify=verify_options) except connection_errors() as e: admin.log_http_error(e, report_url) DOC_PAGE_RE = re.compile("Number of Pages: ([0-9]*),") DOC_CREATION_DATE_RE = re.compile("Create Time/Date: ([A-Za-z 0-9:]*),") DOC_MOD_DATE_RE = re.compile("Last Saved Time/Date: ([A-Za-z 0-9:]*),") DOC_TITLE_RE = re.compile("Title: ([^,]*),") DOC_AUTHOR_RE = re.compile("Author: ([^,]*),") def parse_doc_datetime(raw): if raw.strip() == "": return None my_datetime = None try: my_datetime = datetime.strptime(raw, '%a %b %d %H:%M:%S %Y') except ValueError: pass if my_datetime: return datetime.strftime(my_datetime, '%Y-%m-%d') else: logging.warn('Could not parse DOC date: %s' % raw) return None def metadata_from_doc(doc_path): if not check_tool_present("file", "-v"): logging.warn("Install file to extract metadata! " "The file executable must be in a directory that is in your " "PATH environment variable.") return None real_doc_path = os.path.expandvars(os.path.join(data_dir(), doc_path)) real_doc_path = os.path.abspath(real_doc_path) try: output = subprocess.check_output(["file", real_doc_path], shell=False) output = output.decode('utf-8', errors='replace') except subprocess.CalledProcessError as exc: logging.warn("Error extracting metadata for %s:\n\n%s" % (doc_path, format_exception(exc))) return None metadata = {} page_match = DOC_PAGE_RE.search(output) if page_match: metadata['page_count'] = int(page_match.group(1)) creation_date_match = DOC_CREATION_DATE_RE.search(output) if creation_date_match: metadata['creation_date'] = parse_doc_datetime(creation_date_match.group(1)) mod_date_match = DOC_MOD_DATE_RE.search(output) if mod_date_match: metadata['mod_date'] = parse_doc_datetime(mod_date_match.group(1)) title_match = DOC_TITLE_RE.search(output) if title_match: metadata['title'] = title_match.group(1) author_match = DOC_AUTHOR_RE.search(output) if author_match: metadata['author'] = author_match.group(1) if metadata: return metadata return None def metadata_from_docx(docx_path): try: real_docx_path = os.path.expandvars(os.path.join(data_dir(), docx_path)) real_docx_path = os.path.abspath(real_docx_path) document = docx.Document(real_docx_path) core_props = document.core_properties metadata = {} if core_props.author: metadata['author'] = core_props.author if core_props.title: metadata['title'] = core_props.title if core_props.created: metadata['creation_date'] = datetime.strftime(core_props.created, '%Y-%m-%d') if core_props.modified: metadata['mod_date'] = datetime.strftime(core_props.created, '%Y-%m-%d') if core_props.keywords: metadata['keywords'] = core_props.keywords if metadata: return metadata return None except zipfile.BadZipFile as exc: logging.warn("Error extracting metadata for %s:\n\n%s" % (docx_path, format_exception(exc))) return None def format_exception(exception): exc_type, exc_value, exc_traceback = sys.exc_info() return "\n".join(traceback.format_exception(exc_type, exc_value, exc_traceback)) # assumes working dir is the root dir def data_dir(): if admin.config and admin.config.get('data_directory'): return admin.config.get('data_directory') return "data" def write(content, destination, binary=False): mkdir_p(os.path.dirname(destination)) if binary: f = open(destination, 'bw') else: f = open(destination, 'w', encoding='utf-8') f.write(content) f.close() def json_for(object): return json.dumps(object, sort_keys=True, indent=2, default=format_datetime) def format_datetime(obj): if isinstance(obj, datetime.date): return obj.isoformat() elif isinstance(obj, str): return obj else: return None # mkdir -p in python, from: # http://stackoverflow.com/questions/600268/mkdir-p-functionality-in-python def mkdir_p(path): try: os.makedirs(path) except OSError as exc: # Python >2.5 if exc.errno == errno.EEXIST: pass else: raise # taken from http://effbot.org/zone/re-sub.htm#unescape-html def unescape(text): def remove_unicode_control(str): remove_re = re.compile('[\x00-\x08\x0B-\x0C\x0E-\x1F\x7F]') return remove_re.sub('', str) def fixup(m): text = m.group(0) if text[:2] == "&#": # character reference try: if text[:3] == "&#x": return chr(int(text[3:-1], 16)) else: return chr(int(text[2:-1])) except ValueError: pass else: # named entity try: text = chr(html.entities.name2codepoint[text[1:-1]]) except KeyError: pass return text # leave as is text = re.sub("&#?\w+;", fixup, text) text = remove_unicode_control(text) return text # 'safe' scrapers listed in safe.yml def safe_igs(): return yaml.load(open("safe.yml"))
11492648
from xwing.network.transport.socket.backend.rfc1078 import send, recv class Connection: def __init__(self, loop, sock): self.loop = loop self.sock = sock async def recv(self): '''Try to recv data. If not data is recv NoData exception will raise. :param timeout: Timeout in seconds. `None` meaning forever. ''' return await recv(self.loop, self.sock) async def recv_str(self, encoding='utf-8'): data = await self.recv() if encoding: data = data.decode(encoding) return data async def send(self, data): '''Send data to connected client. :param data: Data to send. ''' return await send(self.loop, self.sock, data) async def send_str(self, data, encoding='utf-8'): if encoding: data = bytes(data, encoding) return await self.send(data) def close(self): self.sock.close()
11492659
import os import uuid import boto3 from six.moves import configparser from botocore.exceptions import ClientError from .aws_util_exceptions import ProfileParsingError from .aws_util_exceptions import RoleNotFoundError from .aws_util_exceptions import AssumeRoleError def get_aws_account_id(profile=None): if profile: session = boto3.Session(profile_name=profile) client = session.client('sts') else: client = boto3.client('sts') account_id = client.get_caller_identity()['Account'] return account_id def get_credential_method_description(session): """Provides a helpful message describing the current IAM execution context.""" profile = '' try: profile = session.profile_name except: pass try: credentials = session.get_credentials() return "{} ({}{})".format( credentials.method, "profile {} -> ".format(profile) if profile != 'default' else '', credentials.access_key ) except: return 'error describing session credentials' def get_boto3_session(aws_creds): if aws_creds: session_token = None if 'AWS_SESSION_TOKEN' in aws_creds: session_token = aws_creds['AWS_SESSION_TOKEN'] return boto3.Session( aws_access_key_id=aws_creds['AWS_ACCESS_KEY_ID'], aws_secret_access_key=aws_creds['AWS_SECRET_ACCESS_KEY'], aws_session_token=session_token, ) else: return boto3.Session() def get_aws_profile_credentials(profile_name, verbose=False): aws_creds = {} # the source_profile may be overridden if a source_profile is indicated in ~/.aws/config source_profile = profile_name config = configparser.ConfigParser() config_file_path = os.path.join(os.path.expanduser("~"),'.aws/config') config.read([config_file_path]) try: section = 'profile {}'.format(profile_name) if not config.has_section(section): msg = "Profile {} not found in {}".format(profile_name, config_file_path) raise ProfileParsingError(msg) if not config.has_option(section, 'role_arn'): if verbose: print("Profile {} in ~/.aws/config does not indicate a role_arn".format(profile_name)) else: aws_creds['role_arn'] = config.get(section, 'role_arn') print("Profile {} indicates role to assume: {}".format(profile_name, aws_creds['role_arn'])) if config.has_option(section, 'source_profile'): source_profile = config.get(section, 'source_profile') print("Found profile {} in ~/.aws/config, indicated source_profile {}".format( profile_name, source_profile )) else: msg = "Profile {} in ~/.aws/config does not indicate a source_profile needed to assume role {}".format( profile_name, aws_creds['role_arn'] ) raise ProfileParsingError(msg) except configparser.ParsingError: print('Error parsing AWS config file') raise except (configparser.NoSectionError, configparser.NoOptionError): print('Error parsing sections or options for AWS profile {} in {}'.format( profile_name, config_file_path)) raise credentials = configparser.ConfigParser() credentials_file_path = os.path.join(os.path.expanduser("~"),'.aws/credentials') credentials.read([credentials_file_path]) try: aws_creds['AWS_ACCESS_KEY_ID'] = credentials.get(source_profile, 'aws_access_key_id') aws_creds['AWS_SECRET_ACCESS_KEY'] = credentials.get(source_profile, 'aws_secret_access_key') try: aws_creds['AWS_SESSION_TOKEN'] = credentials.get(source_profile, 'aws_session_token') except: pass if verbose: print("Found source profile {} in ~/.aws/credentials, access key: {}".format( source_profile, aws_creds['AWS_ACCESS_KEY_ID'] )) except configparser.ParsingError: print('Error parsing AWS credentials file') raise except (configparser.NoSectionError, configparser.NoOptionError): print('Unable to find AWS profile named {} in {}'.format( profile_name, credentials_file_path)) raise return aws_creds def get_role_arn_from_name(aws_creds, role_name, verbose=False): try: session = get_boto3_session(aws_creds) iam_client = session.client('iam') role_arn = iam_client.get_role(RoleName=role_name)['Role']['Arn'] return role_arn except ClientError as e: if verbose: print(e) method = get_credential_method_description(session) if e.response['Error']['Code'] == 'NoSuchEntity': raise RoleNotFoundError(method, e) else: raise AssumeRoleError(method, "Error reading role arn for role name {}: {}".format(role_name, e)) except Exception as e: if verbose: print(e) method = get_credential_method_description(session) raise AssumeRoleError(method, "Error reading role arn for role name {}: {}".format(role_name, e)) def generate_aws_temp_creds(role_arn, aws_creds=None, verbose=False): session = get_boto3_session(aws_creds) sts_client = session.client('sts') aws_creds = {} try: random_session = uuid.uuid4().hex assumed_role_object = sts_client.assume_role( RoleArn=role_arn, RoleSessionName="iamstarter-session-{}".format(random_session), DurationSeconds=3600 # 1 hour max ) aws_creds['AWS_ACCESS_KEY_ID'] = assumed_role_object["Credentials"]["AccessKeyId"] aws_creds['AWS_SECRET_ACCESS_KEY'] = assumed_role_object["Credentials"]["SecretAccessKey"] aws_creds['AWS_SESSION_TOKEN'] = assumed_role_object["Credentials"]["SessionToken"] except Exception as e: if verbose: print(e) method = get_credential_method_description(session) raise AssumeRoleError(method, "Error assuming role {}: {}".format(role_arn, e)) return aws_creds
11492668
from socket import AF_INET, SOCK_DGRAM, socket # import requests import sys host = "172.16.58.3" seq_num = sys.argv[1] method = sys.argv[2] buf = 2048 # url = "http://localhost:8002/getUDPPort" # response = requests.get(url) # data = response.content # port = int(data.decode('utf-8')) port = 20000 addr = (host, port) udp_string = seq_num + "," + method udp_socket = socket(AF_INET, SOCK_DGRAM) udp_socket.sendto(udp_string.encode(), addr) print("Sending %s ..." % udp_string) udp_socket.close()
11492669
import sys sys.path.insert(0, '..') import unittest import json import numpy as np import torch from model import DurIAN from base import suite, BaseModelForwardPassTest seed = 0 np.random.seed(seed) torch.manual_seed(seed) torch.backends.cudnn.deterministic = True class DurIANForwardPassTest(BaseModelForwardPassTest): def __init__(self, *args, **kwargs): super(DurIANForwardPassTest, self).__init__(*args, **kwargs) self.CLASS_TYPE = DurIAN with open('../configs/default.json') as f: self.config = json.load(f) self.config['n_symbols'] = 100 if __name__ == '__main__': runner = unittest.TextTestRunner() runner.run(suite([DurIANForwardPassTest]))
11492695
import copy import random import gym import numpy as np from gym import spaces from gym.utils import seeding class Bit_Flipping_Environment(gym.Env): environment_name = "Bit Flipping Game" def __init__(self, environment_dimension=20, deterministic=False): self.action_space = spaces.Discrete(environment_dimension) self.observation_space = spaces.Dict(dict( desired_goal=spaces.Box(0, 1, shape=(environment_dimension,), dtype='float32'), achieved_goal=spaces.Box(0, 1, shape=(environment_dimension,), dtype='float32'), observation=spaces.Box(0, 1, shape=(environment_dimension,), dtype='float32'), )) self.seed() self.reward_threshold = 0.0 self.trials = 50 self.max_episode_steps = environment_dimension self.id = "Bit Flipping" self.environment_dimension = environment_dimension self.reward_for_achieving_goal = self.environment_dimension self.step_reward_for_not_achieving_goal = -1 self.deterministic = deterministic def seed(self, seed=None): self.np_random, seed = seeding.np_random(seed) return [seed] def reset(self): if not self.deterministic: self.desired_goal = self.randomly_pick_state_or_goal() self.state = self.randomly_pick_state_or_goal() else: self.desired_goal = [0 for _ in range(self.environment_dimension)] self.state = [1 for _ in range(self.environment_dimension)] self.state.extend(self.desired_goal) self.achieved_goal = self.state[:self.environment_dimension] self.step_count = 0 return {"observation": np.array(self.state[:self.environment_dimension]), "desired_goal": np.array(self.desired_goal), "achieved_goal": np.array(self.achieved_goal)} def randomly_pick_state_or_goal(self): return [random.randint(0, 1) for _ in range(self.environment_dimension)] def step(self, action): """Conducts the discrete action chosen and updated next_state, reward and done""" if type(action) is np.ndarray: action = action[0] assert action <= self.environment_dimension + 1, "You picked an invalid action" self.step_count += 1 if action != self.environment_dimension + 1: #otherwise no bit is flipped self.next_state = copy.copy(self.state) self.next_state[action] = (self.next_state[action] + 1) % 2 if self.goal_achieved(self.next_state): self.reward = self.reward_for_achieving_goal self.done = True else: self.reward = self.step_reward_for_not_achieving_goal if self.step_count >= self.environment_dimension: self.done = True else: self.done = False self.achieved_goal = self.next_state[:self.environment_dimension] self.state = self.next_state return {"observation": np.array(self.next_state[:self.environment_dimension]), "desired_goal": np.array(self.desired_goal), "achieved_goal": np.array(self.achieved_goal)}, self.reward, self.done, {} def goal_achieved(self, next_state): return next_state[:self.environment_dimension] == next_state[-self.environment_dimension:] def compute_reward(self, achieved_goal, desired_goal, info): """Computes the reward we would have got with this achieved goal and desired goal. Must be of this exact interface to fit with the open AI gym specifications""" if (achieved_goal == desired_goal).all(): reward = self.reward_for_achieving_goal else: reward = self.step_reward_for_not_achieving_goal return reward
11492702
import pykd import re from common.v_0_0_3.common_utils import * print('='*10 + ' Start ' + '='*10) runCmd(r'bc *;g') #runCmd(r'!idna.tt 4159E80000018') # Abnormal:ieframe!CDownloadWindowItem::_SetState #runCmd(r'!idna.tt 54BF700000644') # Abnormal:wininet!CommitUrlCacheEntryW #runCmd(r'!idna.tt CC66400005FC') # Normal: wininet!CommitUrlCacheEntryW runCmd(r'bp ieframe!CDownloadSecurity::_SendSecurityErrorMessage') runCmd(r'bp ieframe!CDownloadWindowItem::_SetState') runCmd(r'bp ieframe!CNotificationBar2::SetFormattedText') runCmd(r'bp wininet!CommitUrlCacheEntryW') runCmd(r'bp wininet!CCacheServerContainer::AddUrl') runCmd(r'bp wininet!CCacheClientContainer::AddUrl') #runCmd(r'bp rpcrt4!LRPC_BASE_CCALL::SendReceive') # Detail * 0.5 #runCmd(r'bp ntdll!ZwAlpcSendWaitReceivePort') # Detail * 0.5 #runCmd(r'bp rpcrt4!NdrClientCall3') # Detail #runCmd(r'bp rpcrt4!NdrpClientCall3') # Detail #runCmd(r'bp rpcrt4!Ndr64pClientUnMarshal') # Detail #runCmd(r'bp ntdll!memcpy') # Detail *2 runCmd(r'bd *;g-;be *') runCmdLog(r'bl', False) while True: ret = runCmd(r'g') if ttt_test2end(ret): pyLog('='*10 + ' End ' + '='*10) break runCmd(r'.time') runCmdLog(r'kL3', False) #if test2Time('54C4AC0000046'): # Abnormal: Out of wininet!CommitUrlCacheEntryW #if test2Time('CC79C000003A'): # Normal: Out of wininet!CommitUrlCacheEntryW # break ret = runCmd(r'kP') for line in ret.split('\n'): if 'eState = DLState' in line \ or 'wchar_t * pwzOriginDownloadUrl = ' in line \ or 'wchar_t * pwzDestinationFilePath = ' in line \ or 'wchar_t * psz' in line: pyLog(line) #if 'ieframe!CDownloadSecurity::_SendSecurityErrorMessage' in ret: # break ret = runCmd(r'!mex.t') for line in ret.split('\n'): if 'webcache_' in line: pyLog(line) LOG_FILE.flush() LOG_FILE.close()
11492704
import enum from typing import NamedTuple, Optional @enum.unique class ErdAcFanSetting(enum.Enum): DEFAULT = 0 AUTO = 1 LOW = 2 LOW_AUTO = 3 MED = 4 MED_AUTO = 5 HIGH = 8 HIGH_AUTO = 9 def stringify(self, **kwargs): return self.name.replace("_"," ").title() @enum.unique class ErdAcOperationMode(enum.Enum): COOL = 0 FAN_ONLY = 1 ENERGY_SAVER = 2 HEAT = 3 DRY = 4 AUTO = 5 DEFAULT = 9 def stringify(self, **kwargs): return self.name.replace("_"," ").title() @enum.unique class ErdAcFilterStatus(enum.Enum): OK = 0 CLEAN = 1 DEFAULT = -1 def stringify(self, **kwargs): return self.name.replace("_"," ").title() def boolify(self) -> Optional[bool]: return self != ErdAcFilterStatus.OK
11492726
import unittest, sys sys.path.extend(['.','..','../..','py']) import h2o, h2o_common print "Assumes you ran ../build_for_clone.py in this directory" print "Using h2o-nodes.json. Also the sandbox dir" # uses a reduced common class class releaseTest(h2o_common.ReleaseCommon2, unittest.TestCase): def test_shutdown(self): h2o.nodes[0].shutdown_all() if __name__ == '__main__': h2o.unit_main()
11492728
from setuptools import setup, find_packages with open("README.md") as f: long_description = f.read() setup( name="pytest-pylenium", description="selenium wrapper for pytest to aid with system testing", license="Apache Software License 2.0", author="<NAME>", url="https://github.com/symonk/pytest-pylenium", version="0.0.1", author_email="<EMAIL>", maintainer="<NAME>", maintainer_email="<EMAIL>", packages=find_packages(where="src"), package_dir={"": "src"}, entry_points={"pytest11": ["pylenium = pylenium.plugin"]}, setup_requires=["setuptools_scm"], zip_safe=False, classifiers=[ "Development Status :: 4 - Beta", "Framework :: Pytest", "Topic :: Software Development :: Testing", "Topic :: Software Development :: Quality Assurance", "Programming Language :: Python :: 3.7", "Operating System :: OS Independent", "License :: OSI Approved :: Apache Software License", ], )
11492764
from __future__ import print_function from bs4 import BeautifulSoup from urllib.request import urlopen import re import pandas as pd import pickle import threading import sys class FOMC (object): ''' A convenient class for extracting meeting minutes from the FOMC website Example Usage: fomc = FOMC() df = fomc.get_statements() fomc.pickle("./df_minutes.pickle") ''' def __init__(self, base_url='https://www.federalreserve.gov', calendar_url='https://www.federalreserve.gov/monetarypolicy/fomccalendars.htm', historical_date = 2011, verbose = True, max_threads = 10): self.base_url = base_url self.calendar_url = calendar_url self.df = None self.links = None self.dates = None self.articles = None self.verbose = verbose self.HISTORICAL_DATE = historical_date self.MAX_THREADS = max_threads def _get_links(self, from_year): ''' private function that sets all the links for the FOMC meetings from the giving from_year to the current most recent year ''' if self.verbose: print("Getting links...") self.links = [] fomc_meetings_socket = urlopen(self.calendar_url) soup = BeautifulSoup(fomc_meetings_socket, 'html.parser') statements = soup.find_all('a', href=re.compile('^/newsevents/pressreleases/monetary\d{8}a.htm')) self.links = [statement.attrs['href'] for statement in statements] if from_year <= self.HISTORICAL_DATE: for year in range(from_year, self.HISTORICAL_DATE + 1): fomc_yearly_url = self.base_url + '/monetarypolicy/fomchistorical' + str(year) + '.htm' fomc_yearly_socket = urlopen(fomc_yearly_url) soup_yearly = BeautifulSoup(fomc_yearly_socket, 'html.parser') statements_historical = soup_yearly.findAll('a', text = 'Statement') for statement_historical in statements_historical: self.links.append(statement_historical.attrs['href']) def _date_from_link(self, link): date = re.findall('[0-9]{8}', link)[0] if date[4] == '0': date = "{}/{}/{}".format(date[:4], date[5:6], date[6:]) else: date = "{}/{}/{}".format(date[:4], date[4:6], date[6:]) return date def _add_article(self, link, index=None): ''' adds the related article for 1 link into the instance variable index is the index in the article to add to. Due to concurrent prcessing, we need to make sure the articles are stored in the right order ''' if self.verbose: sys.stdout.write(".") sys.stdout.flush() # date of the article content self.dates.append(self._date_from_link(link)) statement_socket = urlopen(self.base_url + link) statement = BeautifulSoup(statement_socket, 'html.parser') paragraphs = statement.findAll('p') self.articles[index]= "\n\n".join([paragraph.get_text().strip() for paragraph in paragraphs]) def _get_articles_multi_threaded(self): ''' gets all articles using multi-threading ''' if self.verbose: print("Getting articles - Multi-threaded...") self.dates, self.articles = [], ['']*len(self.links) jobs = [] # initiate and start threads: index = 0 while index < len(self.links): if len(jobs) < self.MAX_THREADS: t = threading.Thread(target=self._add_article, args=(self.links[index],index,)) jobs.append(t) t.start() index += 1 else: # wait for threads to complete and join them back into the main thread t = jobs.pop(0) t.join() for t in jobs: t.join() for row in range(len(self.articles)): self.articles[row] = self.articles[row].strip() def get_statements(self, from_year=1994): ''' Returns a Pandas DataFrame of meeting minutes with the date as the index uses a date range of from_year to the most current Input from_year is ignored if it is within the last 5 years as this is meant for parsing much older years ''' self._get_links(from_year) print("There are", len(self.links), 'statements') self._get_articles_multi_threaded() self.df = pd.DataFrame(self.articles, index = pd.to_datetime(self.dates)).sort_index() self.df.columns = ['statements'] return self.df def pick_df(self, filename="../data/minutes.pickle"): if filename: if self.verbose: print("Writing to", filename) with open(filename, "wb") as output_file: pickle.dump(self.df, output_file) if __name__ == '__main__': #Example Usage fomc = FOMC() df = fomc.get_statements() fomc.pickle("./df_minutes.pickle")
11492766
import copy import tensorflow as tf import numpy as np from onnx_tf.handlers.backend_handler import BackendHandler from onnx_tf.handlers.handler import onnx_op from onnx_tf.handlers.handler import tf_func @onnx_op("LRN") @tf_func(tf.nn.lrn) class LRN(BackendHandler): @classmethod def _common(cls, node, **kwargs): attrs = copy.deepcopy(node.attrs) alpha = attrs.get("alpha", 1e-4) attrs.setdefault("beta", 0.75) size = attrs["size"] attrs["alpha"] = alpha / size attrs["depth_radius"] = np.floor([(size - 1) / 2.])[0] # TODO: LRN in tf accepts radius # but in ONNX/Caffe accepts diameter. # This could be a problem. return [ cls.make_tensor_from_onnx_node( node, attrs=attrs, c_last_only=True, **kwargs) ] @classmethod def version_1(cls, node, **kwargs): return cls._common(node, **kwargs) @classmethod def version_13(cls, node, **kwargs): return cls._common(node, **kwargs)
11492812
from .._public.artifactstore.api.ArtifactStoreApi import ArtifactStoreApi from .._public.modeldb.api.CommentApi import CommentApi from .._public.modeldb.api.DatasetServiceApi import DatasetServiceApi from .._public.modeldb.api.DatasetVersionServiceApi import DatasetVersionServiceApi from .._public.modeldb.api.ExperimentRunServiceApi import ExperimentRunServiceApi from .._public.modeldb.api.ExperimentServiceApi import ExperimentServiceApi from .._public.modeldb.api.HydratedServiceApi import HydratedServiceApi from .._public.modeldb.api.JobApi import JobApi from .._public.modeldb.api.LineageApi import LineageApi from .._public.modeldb.api.ProjectServiceApi import ProjectServiceApi class ClientSet(object): def __init__(self, client): self.client = client self.artifactStoreService = ArtifactStoreApi(client, "/api/v1/modeldb") self.commentService = CommentApi(client, "/api/v1/modeldb") self.datasetService = DatasetServiceApi(client, "/api/v1/modeldb") self.datasetVersionService = DatasetVersionServiceApi(client, "/api/v1/modeldb") self.experimentRunService = ExperimentRunServiceApi(client, "/api/v1/modeldb") self.experimentService = ExperimentServiceApi(client, "/api/v1/modeldb") self.hydratedService = HydratedServiceApi(client, "/api/v1/modeldb") self.jobService = JobApi(client, "/api/v1/modeldb") self.lineageService = LineageApi(client, "/api/v1/modeldb") self.projectService = ProjectServiceApi(client, "/api/v1/modeldb")
11492818
from ghizmo.commands import lib import os import yaml import urllib.parse from collections import defaultdict _AUTHORS_INFO_FILES = ["authors-info.yml", "authors-info.json", "admin/authors-info.yml", "admin/authors-info.json"] def assemble_authors(config, args): """ Assemble a list of authors as an AUTHORS.md file based on GitHub repo history and a authors-info.{yml,json} file. Supports roles (for each person) and groups of people (leads, contributors, etc.). """ github = config.github repo = config.repo authors_info_filename = None for filename in _AUTHORS_INFO_FILES: if os.path.isfile(filename): authors_info_filename = filename header = None footer = None groups = [] exclude = [] # Assemble roles, keyed by login. roles = {} if authors_info_filename: yield lib.status("Info from: %s" % authors_info_filename) with open(authors_info_filename, "r", encoding="utf-8") as f: info = yaml.safe_load(f) header = info.get("header") footer = info.get("footer") roles = info.get("roles") groups = info.get("groups") exclude = info.get("exclude") yield {"roles": roles, "groups": groups} else: yield lib.status("No roles file") login_to_user = {} for contributor in repo.contributors(): user = github.user(contributor.login) login_to_user[user.login] = user # If any roles are listed but were somehow missing from the contributors return by the API # (for example the commits weren't linked up to the account properly), include them too. contributors_found = {contributor.login for contributor in repo.contributors()} unknown_contributors = [] for login in roles: if login not in contributors_found: user = github.user(login) if user: yield lib.status("Author has a role but is not returned by GitHub as a contributor: %s (%s)" % (login, user)) else: yield lib.status("Author has a role but is not a contributor or a known user: %s [%s]" % (user, type(user))) unknown_contributors.append(login) login_to_user[login] = user yield lib.status("Found %s authors" % len(login_to_user)) yield lib.status("Found without GitHub user info: %s" % unknown_contributors) # Get a list of each group of logins. grouped_authors = [[] for g in groups] default_group = None assigned_logins = set() for count, group in enumerate(groups): yield lib.status("Group %s: %s" % (count, group)) group["number"] = count if "members" in group: members = group["members"] assert isinstance(members, list) lib.status("members %s" % members) for login in members: grouped_authors[count].append(login) assigned_logins.add(login) else: yield lib.status("group has no members: %s %s %s" % (count, group, grouped_authors[count])) default_group = grouped_authors[count] yield {"all_logins": list(login_to_user.keys()), "assigned_logins": list(assigned_logins)} # Put all unassigned logins into last group. assert default_group is not None, "must have a group with no explicit members for unassigned contributors" default_group.extend(set(login_to_user.keys()).difference(assigned_logins)) # Sort each group alphabetically by login. for logins in grouped_authors: logins.sort(key=lambda login: login.lower()) def format_user(login, name): if login and name: return "%s (%s)" % (name, login) elif login: return login else: raise ValueError("Missing login name") commit_tallies = {} for stat in repo.contributor_statistics(): yield lib.status("contrib stat: login '%s' total '%s'" % (stat.author.login, stat.total)) commit_tallies[stat.author.login] = stat.total yield lib.status("Read %s contributor stats" % len(commit_tallies)) issue_tallies = defaultdict(int) for issue in repo.issues(state="all"): issue_tallies[issue.user.login] += 1 yield lib.status("Read %s issues/PRs" % len(issue_tallies)) yield {"commit_tallies": commit_tallies, "issue_tallies": issue_tallies} with open("AUTHORS.md", "w", encoding="utf-8") as f: f.write("# Authors\n\n") if header: f.write("%s\n\n" % header) for group_number, logins in enumerate(grouped_authors): f.write("\n*%s*\n\n" % groups[group_number]["name"]) for login in logins: user = login_to_user.get(login) role = roles.get(login) name = user.name if user else None if login in exclude: continue user_url = "https://github.com/%s" % login if user else None # Link to commits by that author commits_count = commit_tallies.get(login, 0) commits_url = "%s/commits?author=%s" % (repo.html_url, urllib.parse.quote_plus(login)) # Link to issues and PRs by that author. issues_count = issue_tallies.get(login, 0) issues_url = "%s/issues?q=%s" % (repo.html_url, urllib.parse.quote_plus("author:%s" % login)) yield lib.status("login '%s' commits %s issues %s" % (login, commits_count, issues_count)) user_link = format_user(login, name) if user_url: user_link = "[%s](%s)" % (user_link, user_url) f.write("* %s" % user_link) if commits_count or issues_count: f.write(" — [%s+](%s)/[%s+](%s)" % (commits_count, commits_url, issues_count, issues_url)) if role: f.write(" — _%s_" % role) f.write("\n") if footer: f.write("\n%s\n\n" % footer) f.write("\n(This file was auto-generated by [ghizmo assemble-authors](https://github.com/jlevy/ghizmo).)")
11492819
from typing import Any from typing import Union import numpy as np __all__ = ["Face", "Integer", "Float", "Complex", "Bool", "Void"] Face = Any _NumpyInt = Union[ np.int_, np.intc, np.intp, np.int8, np.int16, np.int32, np.int64, np.uint8, np.uint16, np.uint32, np.uint64, ] _NumpyFloat = Union[np.float_, np.float16, np.float32, np.float64] _NumpyComplex = Union[np.complex_, np.complex64, np.complex128] Integer = Union[int, _NumpyInt] Float = Union[float, _NumpyFloat] Complex = Union[complex, _NumpyComplex] Bool = Union[bool, np.bool_] Void = Union[None, np.void]
11492853
import pandas as pd def compute_quotient_metrics(filename, index_col=0, resample_period='1M', shift=0, quotient_metrics=[('Volume', 'max', 'median'), ('Close', 'max', 'min')]): """Compute quotient metrics from CSV data. Quotient metrices are given in the form: (Name, Num, Den), where Name is the corresponding column from the CSV file to be measured, Num is the metric to be used as the numerator, and Den is the metric to be used as the denominator. The column specified by index_col is assumed to contain the dates, which are then binned each metric over the resample period. By default, the metrics are computed for the same time period, but this can be changed by setting the shift value. For example, setting shift=1 will cause the metric computed for February to be divided by the metric computed in January. For example: compute_quotient_metrics(filename, index_col=0, resample_period='1M', shift=0, quotient_metrics = [('Volume', 'max', 'median'), ('Close', 'max', 'min')]): computes for each month in the data, the maximum volume divided by the median volume, and the maximum close price over the minimum close price. The output metrics can be further reduced by computing the maximum monthly metric. See pandas.series.resample for examples of valid resample periods and metrics. """ data = pd.read_csv(filename, index_col=index_col, parse_dates=True) def compute_quotient_metric(name, num_metric, den_metric): series = pd.TimeSeries(data[name]) num_period = series.resample(resample_period, how=num_metric) den_period = series.resample(resample_period, how=den_metric) return num_period[shift:]/den_period[:len(num_period)-shift].values return [compute_quotient_metric(*qm) for qm in quotient_metrics] def compute_quotient_metric(data, resample_period, shift, name, num_metric, den_metric): series = pd.TimeSeries(data[name]) num_period = series.resample(resample_period, how=num_metric) den_period = series.resample(resample_period, how=den_metric) return num_period[shift:]/den_period[:len(num_period)-shift].values if __name__ == '__main__': v, p = compute_quotient_metrics('plpl.csv', shift=1) print "Maximum Monthly Volume Max/Median for PLPL : %6.6g" % v.max() print "Maximum Monthly Close Price Max/Min for PLPL: %6.6g" % p.max()
11492872
from datetime import datetime class Field(object): to_py = lambda self, value: value to_linode = to_py def __init__(self, field): self.field = field class IntField(Field): def to_py(self, value): if value is not None and value != '': return int(value) to_linode = to_py class FloatField(Field): def to_py(self, value): if value is not None: return float(value) to_linode = to_py class CharField(Field): to_py = lambda self, value: str(value) to_linode = to_py class BoolField(Field): def to_py(self, value): if value in (1, '1'): return True else: return False def to_linode(self, value): if value: return 1 else: return 0 class ChoiceField(Field): to_py = lambda self, value: value def __init__(self, field, choices=[]): Field.__init__(self, field) self.choices = choices def to_linode(self, value): if value in self.choices: return value else: raise AttributeError class ListField(Field): def __init__(self, field, type=Field(''), delim=','): Field.__init__(self, field) self.__type=type self.__delim=delim def to_linode(self, value): return self.__delim.join([str(self.__type.to_linode(v)) for v in value]) def to_py(self, value): return [self.__type.to_py(v) for v in value.split(self.__delim) if v != ''] class DateTimeField(Field): to_py = lambda self, value: datetime.strptime(value, '%Y-%m-%d %H:%M:%S.0') to_linode = lambda self, value: value.strftime('%Y-%m-%d %H:%M:%S.0') class ForeignField(Field): def __init__(self, field): self.field = field.primary_key self.__model = field def to_py(self, value): return self.__model.get(id=value) def to_linode(self, value): if isinstance(value, int): return value else: return value.id
11492919
import os import numpy as np import PIL import cv2 import tifffile from scipy.signal import convolve2d import merlin from merlin.core import dataset from merlin.data import codebook as cb class MERFISHDataFactory(object): """ A class for simulating MERFISH data sets. """ def __init__(self): self.codebookPath = 'L26E1.csv' self.psfSigma = 1.2 self.imageSize = np.array([1024, 1024]) self.upsampleFactor = 10 self.fluorophoreBrightness = 1000 self.fiducialBrightness = 10000 self.background = 100 self.bitOrganization = [[0, 1], [0, 0], [1, 0], [1, 1], [2, 1], [2, 0], [3, 1], [3, 0], [4, 0], [4, 1], [5, 1], [5, 0], [6, 1], [6, 0], [7, 0], [7, 1]] def simulate_image(self, spotPositions: np.ndarray=None, addNoise: bool=False) -> np.ndarray: """Simulate a single image consisting of point sources with a Gaussian point spread function Args: spotPositions: a n x 2 numpy array containing the positions to simulate the point sources. If not specified, 1000 random positions are selected. addNoise: flag indicating whether poisson noise should be added to the simulated image. Returns: the simulated image """ if spotPositions is None: spotPositions = np.random.uniform(size=(1000, 2)) spotPositions[:, 0] *= self.imageSize[0] spotPositions[:, 1] *= self.imageSize[1] upsampledImage = np.zeros(self.upsampleFactor*self.imageSize) for p in spotPositions: upsampledImage[int(np.floor(p[0]*self.upsampleFactor)), int(np.floor(p[1]*self.upsampleFactor))] += 1000 return self._downsample_image_stack([upsampledImage], addNoise=addNoise)[0] def simulate_dataset(self, datasetName, abundanceScale=1, fluorophoreCount=5, fovCount=10): """Simulate a full MERFISH dataset""" dataDir = os.sep.join([merlin.DATA_HOME, datasetName]) if not os.path.exists(dataDir): os.mkdir(dataDir) simDataset = dataset.DataSet(datasetName) codebook = cb.Codebook(simDataset, self.codebookPath) barcodeNumber = codebook.get_barcode_count() barcodeAbundances = abundanceScale*np.array( [10**np.random.uniform(3) for i in range(barcodeNumber)]) barcodeAbundances[:10] = 0 for i in range(fovCount): merfishImages, rnaPositions = self._simulate_single_fov( codebook, barcodeAbundances, fluorophoreCount) fiducialImage = self._simulate_fiducial_image() tifffile.imsave( os.sep.join([dataDir, 'full_stack_' + str(i) + '.tiff']), merfishImages.astype(np.uint16)) imageCount = np.max([x[0] for x in self.bitOrganization]) + 1 for j in range(imageCount): fileName = 'Conventional_750_650_561_488_405_' + str(i) + \ '_' + str(j) + '.tiff' filePath = os.sep.join([dataDir, fileName]) imageData = np.zeros( shape=(5, *self.imageSize), dtype=np.uint16) firstBitIndex = [i for i,x in enumerate(self.bitOrganization) \ if x[0] == j and x[1] == 0][0] secondBitIndex = [i for i,x in enumerate(self.bitOrganization) \ if x[0] == j and x[1] == 1][0] imageData[0,:,:] = merfishImages[firstBitIndex] imageData[1,:,:] = merfishImages[secondBitIndex] imageData[2,:,:] = fiducialImage tifffile.imsave(filePath, imageData) np.save(os.sep.join( [dataDir, 'true_positions_' + str(i) + '.npy']), rnaPositions) def _simulate_fiducial_image(self): fiducialPositions = np.random.uniform(size=(1000,2)) upsampledFiducials = self.fiducialBrightness*np.histogram2d( fiducialPositions[:,0]*self.imageSize[0], fiducialPositions[:,1]*self.imageSize[1], bins=self.upsampleFactor*self.imageSize)[0] return self._downsample_image_stack([upsampledFiducials])[0] def _simulate_single_fov(self, codebook, barcodeAbundances, fluorophoreCount): barcodeCount = len(barcodeAbundances) bitNumber = codebook.get_bit_count() imageSize = self.imageSize rnaCounts = np.random.poisson(barcodeAbundances) rnaPositions = [np.random.uniform(size=(c, 2)) for c in rnaCounts] for b in range(barcodeCount): rnaPositions[b][:, 0] *= imageSize[0] rnaPositions[b][:, 1] *= imageSize[1] upsampledStack = np.zeros((bitNumber, *self.upsampleFactor*imageSize)) for b in range(barcodeCount): self._add_spots_for_barcode( codebook.get_barcode(b), rnaPositions[b], fluorophoreCount, upsampledStack) imageStack = self._downsample_image_stack(upsampledStack) return imageStack, rnaPositions def _add_spots_for_barcode(self, barcode, positions, fluorophoreCount, upsampledStack): upsampledImage = np.zeros(self.upsampleFactor*self.imageSize) for p in positions: upsampledImage[int(np.floor(p[0]*self.upsampleFactor)), \ int(np.floor(p[1]*self.upsampleFactor))] += 1 upsampledImage = self.fluorophoreBrightness*np.random.poisson( upsampledImage*fluorophoreCount) for i in np.where(barcode)[0]: np.add(upsampledStack[i], upsampledImage, out=upsampledStack[i]) def _downsample_image_stack(self, upsampledStack, addNoise=True): imageStack = np.zeros((len(upsampledStack), *self.imageSize)) for i in range(len(imageStack)): blurredImage = cv2.GaussianBlur(upsampledStack[i].astype(float), ksize=(51, 51), sigmaX=self.upsampleFactor*self.psfSigma) downsampledImage = np.array(PIL.Image.fromarray( convolve2d(blurredImage, np.ones((self.upsampleFactor, self.upsampleFactor))))\ .resize(self.imageSize, PIL.Image.BILINEAR)) if addNoise: imageStack[i] = np.random.poisson( downsampledImage + self.background) else: imageStack[i] = downsampledImage + self.background return imageStack
11492939
def helper(lst1,lst2,n,x): lst1.sort(reverse=True) for i in range(0,n): if(lst1[i] + lst2[i] <= x): continue else: return False return True t = int(input()) while t > 0: n,x = map(int,input().split()) lst1 = list(map(int,input().split())) lst2 = list(map(int,input().split())) if(helper(lst1,lst2,n,x)): print("YES") else: print("NO") t-=1
11492957
import pygame as pg from Entity import Entity from Const import * class Goombas(Entity): def __init__(self, x_pos, y_pos, move_direction): super().__init__() self.rect = pg.Rect(x_pos, y_pos, 32, 32) if move_direction: self.x_vel = 1 else: self.x_vel = -1 self.crushed = False self.current_image = 0 self.image_tick = 0 self.images = [ pg.image.load('images/goombas_0.png').convert_alpha(), pg.image.load('images/goombas_1.png').convert_alpha(), pg.image.load('images/goombas_dead.png').convert_alpha() ] self.images.append(pg.transform.flip(self.images[0], 0, 180)) def die(self, core, instantly, crushed): if not instantly: core.get_map().get_player().add_score(core.get_map().score_for_killing_mob) core.get_map().spawn_score_text(self.rect.x + 16, self.rect.y) if crushed: self.crushed = True self.image_tick = 0 self.current_image = 2 self.state = -1 core.get_sound().play('kill_mob', 0, 0.5) self.collision = False else: self.y_vel = -4 self.current_image = 3 core.get_sound().play('shot', 0, 0.5) self.state = -1 self.collision = False else: core.get_map().get_mobs().remove(self) def check_collision_with_player(self, core): if self.collision: if self.rect.colliderect(core.get_map().get_player().rect): if self.state != -1: if core.get_map().get_player().y_vel > 0: self.die(core, instantly=False, crushed=True) core.get_map().get_player().reset_jump() core.get_map().get_player().jump_on_mob() else: if not core.get_map().get_player().unkillable: core.get_map().get_player().set_powerlvl(0, core) def update_image(self): self.image_tick += 1 if self.image_tick == 14: self.current_image = 1 elif self.image_tick == 28: self.current_image = 0 self.image_tick = 0 def update(self, core): if self.state == 0: self.update_image() if not self.on_ground: self.y_vel += GRAVITY blocks = core.get_map().get_blocks_for_collision(int(self.rect.x // 32), int(self.rect.y // 32)) self.update_x_pos(blocks) self.update_y_pos(blocks) self.check_map_borders(core) elif self.state == -1: if self.crushed: self.image_tick += 1 if self.image_tick == 50: core.get_map().get_mobs().remove(self) else: self.y_vel += GRAVITY self.rect.y += self.y_vel self.check_map_borders(core) def render(self, core): core.screen.blit(self.images[self.current_image], core.get_map().get_camera().apply(self))
11492987
import sys, pytest sys.path.append("../") LIST_ONE = ["a", "b", "c", "d", "e"] LIST_TWO = ["v", "d", "s", "t", "z"] print(LIST_ONE) print(LIST_TWO) from tkinter import OptionMenu from appJar import gui def get(btn): a = app.getOptionBoxWidget("l1") b = app.getOptionBoxWidget("l2") c = app.getOptionBoxWidget("l3") print("VAR ARRAY: ", app.n_optionVars["l1"], app.n_optionVars["l2"], app.n_optionVars["l2"]) print("LINKED VARS: ", a.var, b.var, c.var) print("ORIG GET: ", app.getOptionBox("l1"), app.getOptionBox("l2")) print("LINKED VARS: ", a.var.get(), b.var.get()) print("STORED VARS: ", app.n_optionVars["l1"].get(), app.n_optionVars["l2"].get()) def test0(btn=None): print(LIST_ONE[0], LIST_TWO[0]) assert app.getOptionBox("l1") == LIST_ONE[0] assert app.getOptionBox("l2") == LIST_TWO[0] def test1(btn=None): print(LIST_ONE[0], LIST_TWO[0]) obs = app.getAllOptionBoxes() assert obs["l1"] == LIST_ONE[0] assert obs["l2"] == LIST_TWO[0] def test2(btn=None): # select new items - by position app.setOptionBox("l1", 3) app.setOptionBox("l2", 2) print( LIST_ONE[3], LIST_TWO[2]) assert app.getOptionBox("l1") == LIST_ONE[3] assert app.getOptionBox("l2") == LIST_TWO[2] def test3(btn=None): app.clearOptionBox("l1") print( LIST_ONE[0], LIST_TWO[2]) assert app.getOptionBox("l1") == LIST_ONE[0] assert app.getOptionBox("l2") == LIST_TWO[2] def test4(btn=None): app.setOptionBox("l1", 2) app.clearAllOptionBoxes() print( LIST_ONE[0], LIST_TWO[0]) assert app.getOptionBox("l1") == LIST_ONE[0] assert app.getOptionBox("l2") == LIST_TWO[0] def test5(btn=None): # select new items - by position app.setOptionBox("l1", 2) app.setOptionBox("l2", 3) def test6(btn=None): # select new items - by value app.setOptionBox("l1", LIST_ONE[3]) app.setOptionBox("l2", LIST_TWO[1]) app.renameOptionBoxItem("l2", LIST_TWO[0], "newName") assert app.getOptionBox("l1") == LIST_ONE[3] with gui() as app: print("\tTesting options") # add two option boxes assert isinstance(app.addOptionBox("l1", LIST_ONE), OptionMenu) app.addOptionBox("l2", LIST_TWO) with pytest.raises(Exception): app.addOptionBox("l2", LIST_TWO) app.addOptionBox("l3", LIST_TWO) app.addButtons(["0", "1", "2", "3", "4", "5", "6"], [test0, test1, test2, test3, test4, test5, test6]) app.addButton("get", get)
11493020
import unittest import numpy as np import scipy as sp import matplotlib.pyplot as plt from PySeismoSoil.class_ground_motion import Ground_Motion as GM from PySeismoSoil.class_Vs_profile import Vs_Profile from PySeismoSoil.class_frequency_spectrum import Frequency_Spectrum import os from os.path import join as _join f_dir = _join(os.path.dirname(os.path.realpath(__file__)), 'files') class Test_Class_Ground_Motion(unittest.TestCase): def test_loading_data__two_columns_from_file(self): # Two columns from file gm = GM(_join(f_dir, 'sample_accel.txt'), unit='gal') PGA_benchmark = 294.30 # unit: cm/s/s PGV_benchmark = 31.46 # unit: cm/s PGD_benchmark = 38.77 # unit: cm tol = 1e-2 self.assertAlmostEqual(gm.pga_in_gal, PGA_benchmark, delta=tol) self.assertAlmostEqual(gm.pgv_in_cm_s, PGV_benchmark, delta=tol) self.assertAlmostEqual(gm.pgd_in_cm, PGD_benchmark, delta=tol) self.assertAlmostEqual(gm.peak_Arias_Intensity, 1.524, delta=tol) self.assertAlmostEqual(gm.rms_accel, 0.4645, delta=tol) def test_loading_data__two_columns_from_numpy_array(self): # Two columns from numpy array gm = GM(np.array([[0.1, 0.2, 0.3, 0.4], [1, 2, 3, 4]]).T, unit='m/s/s') self.assertAlmostEqual(gm.pga, 4) def test_loading_data__one_column_from_file(self): # One column from file gm = GM(_join(f_dir, 'one_column_data_example.txt'), unit='g', dt=0.2) self.assertAlmostEqual(gm.pga_in_g, 12.0) def test_loading_data__one_column_from_numpy_array(self): # One column from numpy array gm = GM(np.array([1, 2, 3, 4, 5]), unit='gal', dt=0.1) self.assertAlmostEqual(gm.pga_in_gal, 5.0) def test_loading_data__one_column_without_specifying_dt(self): # One column without specifying dt error_msg = 'is needed for one-column `data`.' with self.assertRaisesRegex(ValueError, error_msg): gm = GM(np.array([1, 2, 3, 4, 5]), unit='gal') def test_loading_data__test_invalid_unit_names(self): # Test invalid unit names with self.assertRaisesRegex(ValueError, 'Invalid `unit` name.'): GM(np.array([1, 2, 3, 4, 5]), unit='test', dt=0.1) with self.assertRaisesRegex(ValueError, r"use '/s/s' instead of 's\^2'"): GM(np.array([1, 2, 3, 4, 5]), unit='m/s^2', dt=0.1) def test_differentiation(self): veloc = np.array([[.1, .2, .3, .4, .5, .6], [1, 3, 7, -1, -3, 5]]).T gm = GM(veloc, unit='m', motion_type='veloc') accel_benchmark = np.array( [[.1, .2, .3, .4, .5, .6], [0, 20, 40, -80, -20, 80]] ).T self.assertTrue(np.allclose(gm.accel, accel_benchmark)) def test_integration__artificial_example(self): gm = GM(_join(f_dir, 'two_column_data_example.txt'), unit='m/s/s') v_bench = np.array([[0.1000, 0.1000], # from MATLAB [0.2000, 0.3000], [0.3000, 0.6000], [0.4000, 1.0000], [0.5000, 1.5000], [0.6000, 1.7000], [0.7000, 2.0000], [0.8000, 2.4000], [0.9000, 2.9000], [1.0000, 3.5000], [1.1000, 3.8000], [1.2000, 4.2000], [1.3000, 4.7000], [1.4000, 5.3000], [1.5000, 6.0000]]) u_bench = np.array([[0.1000, 0.0100], # from MATLAB [0.2000, 0.0400], [0.3000, 0.1000], [0.4000, 0.2000], [0.5000, 0.3500], [0.6000, 0.5200], [0.7000, 0.7200], [0.8000, 0.9600], [0.9000, 1.2500], [1.0000, 1.6000], [1.1000, 1.9800], [1.2000, 2.4000], [1.3000, 2.8700], [1.4000, 3.4000], [1.5000, 4.0000]]) self.assertTrue(np.allclose(gm.veloc, v_bench)) self.assertTrue(np.allclose(gm.displ, u_bench)) def test_integration__real_world_example(self): # Note: In this test, the result by cumulative trapezoidal numerical # integration is used as the benchmark. Since it is infeasible to # achieve perfect "alignment" between the two time histories, # we check the correlation coefficient instead of element-wise # check. veloc_ = np.genfromtxt(_join(f_dir, 'sample_accel.txt')) gm = GM(veloc_, unit='m/s', motion_type='veloc') displ = gm.displ[:, 1] displ_cumtrapz = np.append(0, sp.integrate.cumtrapz(veloc_[:, 1], dx=gm.dt)) r = np.corrcoef(displ_cumtrapz, displ)[1, 1] # cross-correlation self.assertTrue(r >= 0.999) def test_fourier_transform(self): gm = GM(_join(f_dir, 'two_column_data_example.txt'), unit='m/s/s') freq, spec = gm.get_Fourier_spectrum(real_val=False).raw_data.T freq_bench = [ 0.6667, 1.3333, 2.0000, 2.6667, 3.3333, 4.0000, 4.6667, 5.3333, ] FS_bench = [ 60.0000 + 0.0000j, -1.5000 + 7.0569j, -1.5000 + 3.3691j, -7.5000 +10.3229j, -1.5000 + 1.3506j, -1.5000 + 0.8660j, -7.5000 + 2.4369j, -1.5000 + 0.1577j, ] self.assertTrue(np.allclose(freq, freq_bench, atol=0.0001, rtol=0.0)) self.assertTrue(np.allclose(spec, FS_bench, atol=0.0001, rtol=0.0)) def test_baseline_correction(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m/s/s') corrected = gm.baseline_correct(show_fig=True) self.assertTrue(isinstance(corrected, GM)) def test_high_pass_filter(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m') hp = gm.highpass(cutoff_freq=1.0, show_fig=True) self.assertTrue(isinstance(hp, GM)) def test_low_pass_filter(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m') lp = gm.lowpass(cutoff_freq=1.0, show_fig=True) self.assertTrue(isinstance(lp, GM)) def test_band_pass_filter(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m') bp = gm.bandpass(cutoff_freq=[0.5, 8], show_fig=True) self.assertTrue(isinstance(bp, GM)) def test_band_stop_filter(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m') bs = gm.bandstop(cutoff_freq=[0.5, 8], show_fig=True) self.assertTrue(isinstance(bs, GM)) def test_amplify_via_profile(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m') vs_prof = Vs_Profile(_join(f_dir, 'profile_FKSH14.txt')) output_motion = gm.amplify(vs_prof, boundary='elastic') self.assertTrue(isinstance(output_motion, GM)) def test_deconvolution(self): # Assert `deconvolve()` & `amplify()` are reverse operations to each other. gm = GM(_join(f_dir, 'sample_accel.txt'), unit='m') vs_prof = Vs_Profile(_join(f_dir, 'profile_FKSH14.txt')) for boundary in ['elastic', 'rigid']: deconv_motion = gm.deconvolve(vs_prof, boundary=boundary) output_motion = deconv_motion.amplify(vs_prof, boundary=boundary) self.assertTrue(self.nearly_identical(gm.accel, output_motion.accel)) amplified_motion = gm.amplify(vs_prof, boundary=boundary) output_motion = amplified_motion.deconvolve(vs_prof, boundary=boundary) self.assertTrue(self.nearly_identical(gm.accel, output_motion.accel)) def test_plot(self): filename = _join(f_dir, 'sample_accel.txt') gm = GM(filename, unit='m') fig, axes = gm.plot() # automatically generate fig/ax objects self.assertTrue(isinstance(axes, tuple)) self.assertEqual(len(axes), 3) self.assertEqual(axes[0].title.get_text(), os.path.split(filename)[1]) fig2 = plt.figure(figsize=(8, 8)) fig2_, axes = gm.plot(fig=fig2) # feed an external figure object self.assertTrue(np.allclose(fig2_.get_size_inches(), (8, 8))) def test_unit_convert(self): data = np.array([1, 3, 7, -2, -10, 0]) gm = GM(data, unit='m', dt=0.1) accel = gm.accel[:, 1] accel_in_m = gm._unit_convert(unit='m/s/s')[:, 1] accel_in_gal = gm._unit_convert(unit='gal')[:, 1] accel_in_g = gm._unit_convert(unit='g')[:, 1] self.assertTrue(np.allclose(accel_in_m, accel)) self.assertTrue(np.allclose(accel_in_gal, accel * 100)) self.assertTrue(np.allclose(accel_in_g, accel / 9.81)) def test_scale_motion(self): data = np.array([1, 3, 7, -2, -10, 0]) gm = GM(data, unit='g', dt=0.1) gm_scaled_1 = gm.scale_motion(factor=2.0) # scale by 2.0 gm_scaled_2 = gm.scale_motion(target_PGA_in_g=5.0) # scale by 0.5 self.assertTrue(np.allclose(gm.accel[:, 1] * 2, gm_scaled_1.accel[:, 1])) self.assertTrue(np.allclose(gm.accel[:, 1] * 0.5, gm_scaled_2.accel[:, 1])) def test_amplify_by_tf__case_1_an_artificial_transfer_function(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='gal') ratio_benchmark = 2.76 freq = np.arange(0.01, 50, step=0.01) tf = ratio_benchmark * np.ones_like(freq) transfer_function = Frequency_Spectrum(np.column_stack((freq, tf))) new_gm = gm.amplify_by_tf(transfer_function, show_fig=False) ratio = new_gm.accel[:, 1] / gm.accel[:, 1] self.assertTrue(np.allclose(ratio, ratio_benchmark)) def test_amplify_by_tf__case_2_a_transfer_function_from_a_Vs_profile(self): gm = GM(_join(f_dir, 'sample_accel.txt'), unit='gal') vs_prof = Vs_Profile(_join(f_dir, 'profile_FKSH14.txt')) tf_RO, tf_BH, _ = vs_prof.get_transfer_function() gm_with_tf_RO = gm.amplify_by_tf(tf_RO) gm_with_tf_BH = gm.amplify_by_tf(tf_BH) gm_with_tf_RO_ = gm.amplify(vs_prof, boundary='elastic') gm_with_tf_BH_ = gm.amplify(vs_prof, boundary='rigid') # Assert that `amplify_by_tf()` and `amplify()` can generate # nearly identical results self.assertTrue( self.nearly_identical(gm_with_tf_RO.accel, gm_with_tf_RO_.accel) ) self.assertTrue( self.nearly_identical(gm_with_tf_BH.accel, gm_with_tf_BH_.accel) ) @staticmethod def nearly_identical(motion_1, motion_2, thres=0.99): """ Assert that two ground motions are nearly identical, by checking the correlation coefficient between two time series. Parameters ---------- motion_1 : numpy.ndarray Two-column array (time, acceleration). motion_2 : numpy.ndarray Two-column array (time, acceleration). thres : float The threshold that the correlation coefficient must be above (or equal to). Returns ------- result : bool Whether the motions are nearly identical """ if not np.allclose(motion_1[:, 0], motion_2[:, 0], rtol=0.001, atol=0.0): return False r = np.corrcoef(motion_1[:, 1], motion_2[:, 1]) if r[1, 0] < thres: return False return True if __name__ == '__main__': SUITE = unittest.TestLoader().loadTestsFromTestCase(Test_Class_Ground_Motion) unittest.TextTestRunner(verbosity=2).run(SUITE)
11493041
from toee import * import char_editor def CheckPrereq(attachee, classLevelled, abilityScoreRaised): #Str requirement enforced by the engine if not char_editor.has_feat(feat_two_weapon_fighting) and not char_editor.has_feat(feat_two_weapon_fighting_ranger): return 0 return 1
11493045
from .AbstractAdjacencyCompander import AbstractAdjacencyCompander import numpy as np import sys import os import os.path sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__),'../../../../preprocessing'))) from mnist_to_graph_tensor import mnist_adj_mat class ImageAdjacencyCompander(AbstractAdjacencyCompander): def __init__(self,V,A): super(ImageAdjacencyCompander, self).__init__(V, A) self.NUM_DIRS = 8 def contractA(self): self.flatA = self.A.sum(axis=1) return self.flatA def expandA(self): return self.A def update(self, P): Ptiled = np.tile(np.expand_dims(P,axis=1),(1,self.NUM_DIRS,1)) Ptranspose = np.transpose(Ptiled, axes=[1, 2, 0]) Pnottranspose = np.transpose(Ptiled, axes=[1, 0, 2]) Abatched = np.transpose(self.A, axes=[1, 0, 2]) leftMultiply = np.matmul(Ptranspose, Abatched) rightMultiply = np.matmul(leftMultiply, Pnottranspose) self.A = np.transpose(rightMultiply, axes=[1, 0, 2]) self.V = np.dot(P.transpose(), self.V) self.N = self.V.shape[0]
11493048
import os from infiniteremixer.utils.io import load class BatchExtractor: """Batch extract features dynamically from a list of audio files.""" def __init__(self, sample_rate=22050): self.sample_rate = sample_rate self.extractors = [] self._features = {} def add_extractor(self, extractor): """Add a concrete Extractor to the extractors. :param extractor: (Extractor) Concrete Extractor (e.g., MFCCExtractor) """ self.extractors.append(extractor) def extract(self, dir): """Extract features from all the audio files in the directory. :param dir: (str) Path to directory with audio files to analyse :return: (dict): Dictionary with audio file paths as keys and extracted features in the form: { "filepath1": { "chromogram": np.ndarray([[], [], ...]), "mfcc": np.ndarray([[], [], ...]) }, "filepath2": { "chromogram": np.ndarray([[], [], ...]), "mfcc": np.ndarray([[], [], ...]) }, ... } """ features = {} for root, _, files in os.walk(dir): for file in files: file_path = os.path.join(root, file) file_features = self._extract_features_for_file(file_path) features[file_path] = file_features return features def _extract_features_for_file(self, file_path): features = {} signal = load(file_path, self.sample_rate) for extractor in self.extractors: feature = extractor.extract(signal, self.sample_rate) features[extractor.feature_name] = feature return features if __name__ == "__main__": from infiniteremixer.data.extraction.mfccextractor import MFCCExtractor from infiniteremixer.data.extraction.chromogramextractor import ChromogramExtractor num_mfccs = 13 frame_size = 1024 hop_size = 512 dir = "/home/valerio/datasets/infinitemixer/beats" mfcc_extractor = MFCCExtractor(frame_size, hop_size, num_mfccs) chromogram_extractor = ChromogramExtractor(frame_size, hop_size) batch_extractor = BatchExtractor(22050) batch_extractor.add_extractor(mfcc_extractor) batch_extractor.add_extractor(chromogram_extractor) features = batch_extractor.extract(dir) a = 1
11493136
import numpy as np from scipy.special import jv as besselj from Solvers.QSP_solver import QSP_Solver from math import ceil # -------------------------------------------------------------------------- # Test case 1: Hamiltonian simulation # # Here we want to approxiamte e^{-i\tau x} by Jacobi-Anger expansion: # # e^{-i\tau x} = J_0(\tau)+2\sum_{k even} (-1)^{k/2}J_{k}(\tau)T_k(x)+2i\sum_{k odd} (-1)^{(k-1)/2}J_{k}(\tau) T_k(x) # # We truncate the series up to N = 1.4\tau+log(10^{14}), which gives an polynomial approximation of e^{-i\tau x} with # accuracy 10^{-14}. Besides, we deal with real and imaginary part of the truncated series seperatly and divide them # by a constant factor 2 to enhance stability. # # parameters # tau: the duration \tau in Hamiltonian simulation # criteria: stop criteria, default 1e-12 # plot_phase: whether plot phase factors # # -------------------------------------------------------------------------- # # Reference: <NAME>, <NAME>, <NAME> and <NAME> # Efficient Phase Factor Evaluation in Quantum Signal Processing # # Author: <NAME>, <NAME> # Version 1.0 # Last Update 06/2020 # # -------------------------------------------------------------------------- # setup parameters tau = 1000 criteria = 1e-12 plot_phase = True opts = dict() # -------------------------------------------------------------------------- # find phase factors opts["criteria"] = criteria max_order = ceil(1.4 * tau + np.log(1e14)) if np.mod(max_order, 2) == 1: max_order -= 1 # -------------------------------------------------------------------------- # even part coeff = np.zeros((max_order//2 + 1, 1)) for i in range(len(coeff)): coeff[i] = (-1)**(i) * besselj(2*i, tau) coeff[0] /= 2 [phi1, out1] = QSP_Solver(coeff, 0, opts) print("- Info: \t\tQSP phase factors --- solved by L-BFGS\n") print("- Parity: \t\t%s\n- Degree: \t\t%d\n", "even", max_order) print("- Iteration times: \t%d\n", out1["iter"]) print("- CPU time: \t%.1f s\n", out1["time"]) #-------------------------------------------------------------------------- # odd part coeff = np.zeros((max_order/2 + 1, 1)) for i in range(len(coeff)): coeff[i] = (-1)**(i) * besselj(2*i + 1, tau) [phi2,out2] = QSP_Solver(coeff, 1, opts) #-------------------------------------------------------------------------- # output print("- Info: \t\tQSP phase factors --- solved by L-BFGS\n") print("- Parity: \t\t%s\n- Degree: \t\t%d\n", "odd", max_order + 1) print("- Iteration times: \t%d\n", out2["iter"]) print("- CPU time: \t%.1f s\n", out2["time"]) #-------------------------------------------------------------------------- # plot phase factors ## Won't plot until necessary
11493153
from django.contrib.auth import get_user_model from django.db.models.signals import m2m_changed from django.dispatch import receiver from guardian.shortcuts import assign_perm, remove_perm from grandchallenge.blogs.models import Post @receiver(m2m_changed, sender=Post.authors.through) def update_permissions_on_authors_changed( instance, action, reverse, pk_set, **_ ): if action not in ["post_add", "post_remove", "pre_clear"]: # nothing to do for the other actions return if reverse: users = [instance] if pk_set is None: # When using a _clear action, pk_set is None # https://docs.djangoproject.com/en/2.2/ref/signals/#m2m-changed posts = instance.blog_authors.all() else: posts = Post.objects.filter(pk__in=pk_set) else: posts = Post.objects.get(pk=instance.pk) if pk_set is None: # When using a _clear action, pk_set is None # https://docs.djangoproject.com/en/2.2/ref/signals/#m2m-changed users = instance.authors.all() else: users = get_user_model().objects.filter(pk__in=pk_set) op = assign_perm if "add" in action else remove_perm for user in users: op("change_post", user, posts)
11493156
from django.conf import settings from django.core import validators class BookmarkURLValidator(validators.URLValidator): """ Extends default Django URLValidator and cancels validation if it is disabled in settings. This allows to switch URL validation on/off dynamically which helps with testing """ def __call__(self, value): if settings.LD_DISABLE_URL_VALIDATION: return super().__call__(value)
11493185
import time def initGPIO(): try: with open("/sys/class/gpio/export", "w") as fp: fp.write("477") except: pass try: with open("/sys/class/gpio/export", "w") as fp: fp.write("476") except: pass try: with open("/sys/class/gpio/gpio477/direction", "w") as fp: fp.write("out") except: pass try: with open("/sys/class/gpio/gpio476/direction", "w") as fp: fp.write("out") except: pass def gpioWriteDC(val): try: with open("/sys/class/gpio/gpio477/value", "w") as fp: fp.write(str(val)) except: pass def gpioWriteReset(val): try: with open("/sys/class/gpio/gpio476/value", "w") as fp: fp.write(str(val)) except: pass def gpioDoReset(): gpioWriteReset(1) time.sleep(0.1) gpioWriteReset(0) time.sleep(0.1) gpioWriteReset(1) time.sleep(0.1)
11493202
import pytest from shortcuts import FMT_SHORTCUT, FMT_TOML from shortcuts.cli import _get_format class Test_get_format: @pytest.mark.parametrize('filepath,exp_format', [ ('file.shortcut', FMT_SHORTCUT), ('file.plist', FMT_SHORTCUT), ('file.toml', FMT_TOML), ('https://icloud.com/shortcuts/some-id/', 'url'), ]) def test_for_url(self, filepath, exp_format): assert _get_format(filepath) == exp_format def test_raises(self): with pytest.raises(RuntimeError): _get_format('abc')
11493211
from sentry_relay import DataCategory, SPAN_STATUS_CODE_TO_NAME def test_parse_data_category(): assert DataCategory.parse("default") == DataCategory.DEFAULT assert DataCategory.parse("transaction") == DataCategory.TRANSACTION assert DataCategory.parse("") is None assert DataCategory.parse(None) is None assert DataCategory.parse("something completely different") is None def test_data_category_from_event_type(): assert DataCategory.from_event_type("transaction") == DataCategory.TRANSACTION # Special case! assert DataCategory.from_event_type("default") == DataCategory.ERROR # Anything unknown is coerced to "default", which is ERROR assert DataCategory.from_event_type("") == DataCategory.ERROR assert DataCategory.from_event_type(None) == DataCategory.ERROR def test_data_category_api_name(): assert DataCategory.ERROR.api_name() == "error" def test_data_category_compatibility(): assert 1 == DataCategory.ERROR assert 1 in DataCategory.event_categories() assert DataCategory.ERROR in (0, 1, 2) def test_span_mapping(): # This is a pure regression test to protect against accidental renames. assert SPAN_STATUS_CODE_TO_NAME == { 0: "ok", 1: "cancelled", 2: "unknown", 3: "invalid_argument", 4: "deadline_exceeded", 5: "not_found", 6: "already_exists", 7: "permission_denied", 8: "resource_exhausted", 9: "failed_precondition", 10: "aborted", 11: "out_of_range", 12: "unimplemented", 13: "internal_error", 14: "unavailable", 15: "data_loss", 16: "unauthenticated", }
11493213
import argparse from typing import List import tqdm import canrevan.parsing as parsing import canrevan.utils as utils from canrevan.crawling import Crawler DEFAULT_USER_AGENT_STRING = ( "Mozilla/5.0 (Windows NT 10.0; Win64; x64) " "AppleWebKit/537.36 (KHTML, like Gecko) " "Chrome/87.0.4280.66 " "Safari/537.36" ) def _main(): args = _create_argument_parser().parse_args() # Create a crawler for collecting article urls and news contents. crawler = Crawler( concurrent_tasks=args.max_jobs, num_parsing_processes=args.num_cores, request_headers={"user-agent": args.user_agent}, request_timeout=args.timeout, ) # Collect article urls from navigation pages. nav_urls = _prepare_nav_urls(args) print(f"[*] navigation pages: {len(nav_urls)}") with tqdm.tqdm(nav_urls, desc="[*] collect article urls") as tbar: article_urls = crawler.reduce_to_array( nav_urls, parse_fn=parsing.extract_article_urls, update_fn=tbar.update ) # Flatten the grouped urls and remove duplicates from the array. article_urls = {url for urls in article_urls for url in urls} print(f"[*] total collected articles: {len(article_urls)}") # Crawl news articles from the collected article urls and save the content to the # output file. with tqdm.tqdm(article_urls, desc="[*] crawl news article contents") as tbar: total_contents = crawler.reduce_to_file( article_urls, args.output_path, parse_fn=parsing.parse_article_content, update_fn=tbar.update, ) print( f"[*] finish crawling {total_contents} news articles to " f"[{args.output_path}]" ) def _prepare_nav_urls(args: argparse.Namespace) -> List[str]: return [ f"https://news.naver.com/main/list.nhn?mode=LSD&mid=shm" f"&sid1={category}&date={date}&page={page}" for category in args.category for date in utils.drange(args.start_date, args.end_date, args.skip_days) for page in range(1, args.max_page + 1) ] def _create_argument_parser() -> argparse.ArgumentParser: parser = argparse.ArgumentParser( prog="canrevan", description="crawl naver news articles" ) parser.add_argument( "--output_path", default="articles.txt", help="output file path" ) parser.add_argument( "--category", required=True, nargs="*", type=int, help="list of news article categories", ) parser.add_argument( "--start_date", required=True, help="minimum date of news articles" ) parser.add_argument( "--end_date", required=True, help="maximum date of news articles" ) parser.add_argument( "--skip_days", default=1, type=int, help="number of days to skip from crawling" ) parser.add_argument( "--max_page", default=10, type=int, help="maximum number of pages to navigate" ) parser.add_argument( "--timeout", default=5, type=float, help="timeout for the whole request" ) parser.add_argument( "--max_jobs", default=500, type=int, help="maximum number of concurrent requests", ) parser.add_argument( "--num_cores", default=4, type=int, help="number of multi-processing cores for parsing", ) parser.add_argument( "--user-agent", default=DEFAULT_USER_AGENT_STRING, help="use custom user-agent string", ) return parser
11493241
import pandas as pd from tqdm import tqdm import json import torch from nltk.corpus import stopwords from transformers import GPT2LMHeadModel from model.ConditionalLM import ConditionalLM from util.utils import load_args, compute_logProb, init_tokenizer from util.CLMDataset import CLMDataset from torch.utils.data import DataLoader from collections import Counter import json from tqdm import tqdm import os class LocatingModule(): def __init__(self, config='configs/locating.yaml'): self.args = load_args(config) print(self.args) if not os.path.isdir('output/locating/{}'.format(self.args.dataset)): os.mkdir('output/locating/{}'.format(self.args.dataset)) self.tokenizer = init_tokenizer(self.args.gpt_path) def load_utt_intent(self, fname): df = pd.read_csv(fname) return [(utt, intent) for utt, intent in zip(list(df['utt']), list(df['intent']))] def load_semantic(self, fname): with open(fname) as f: word_score = json.load(f) intent_related_words = {} for intent, scores in word_score.items(): intent_related_words[intent] = [] for score in scores: if score[1] > self.args.threshold: intent_related_words[intent].append(score[0]) return intent_related_words def token_score_to_word_score(self, utt, score): m_score = [] # score = [float(ts) for ts in scores[i].split(' ')[1:]] # ts: token score utt_tokens_with_space = utt.split(' ') with_space_index = 0 utt_tokens_without_space = utt.split() score_index = 0 m_score = [] # merge score for index, token in enumerate(utt_tokens_without_space): if index == 0: length = len(self.tokenizer.tokenize(token)) with_space_index += 1 else: if utt_tokens_with_space[with_space_index] == '': with_space_index += 2 length = len(self.tokenizer.tokenize('one ' + token)) - 1 else: with_space_index += 1 length = len(self.tokenizer.tokenize('one ' + token)) - 1 tmp = 0 for i in range(score_index, score_index + length): tmp += score[i] score_index += length m_score.append(tmp) assert len(m_score) == len(utt_tokens_without_space) # m_scores.append(' '.join([str(ts)[:5] if len(str(ts)) >= 5 else str(ts) + '0' * (5 - len(str(ts))) for ts in m_score])) return m_score @torch.no_grad() def generate_word_score(self): output = 'output/locating/{}/{}.txt'.format(self.args.dataset, self.args.llr_name) delimiter = '##' gpt2 = GPT2LMHeadModel.from_pretrained(self.args.gpt_path, return_dict=True).to(self.args.device) clm = ConditionalLM(self.args.gpu, self.args.dataset, self.args.label_num).to(self.args.device) clm.load_state_dict(torch.load('output/params/{}/{}.pt'.format(self.args.dataset, self.args.cond_name), \ map_location='cuda:{}'.format(self.args.gpu) if self.args.gpu != -1 else 'cpu')) gpt2.eval() clm.eval() train_loader = self.create_loader(self.args.train, False) with open(output, 'w', encoding='utf-8') as f: for X_in, X_out, mask, lengths, y, intents in tqdm(train_loader): logProb_Cond = compute_logProb(X_out, clm(X_in, y)) logProb = compute_logProb(X_out, gpt2(input_ids=X_in, attention_mask=mask).logits) llrs = logProb_Cond - logProb for x, llr, intent, length in zip(X_out, llrs, intents, lengths): f.write("{}{}".format(intent, delimiter)) x = self.tokenizer.batch_decode(x.reshape(1,-1))[0].replace("<|endoftext|>","") x_word = x.split() x_token_llrs = [r.item() for r in llr[:length]] x_word_scores = self.token_score_to_word_score(x, x_token_llrs) assert len(x_word) == len(x_word_scores) f.write(delimiter.join(x_word)) f.write("\n{}".format(delimiter)+ delimiter.join([str(s) for s in x_word_scores])+ "\n") en_stopwords = list(stopwords.words('english')) summary = {} with open(output) as f: for i, line in enumerate(f): line = line.strip() if i % 2 == 0: text = line.split(delimiter) else: score = list(map(float, line.split(delimiter)[1:])) if text[0] not in summary: summary[text[0]]={} for j, w in enumerate(text[1:]): if w in en_stopwords: # not necessary continue if w not in summary[text[0]]: summary[text[0]][w]=score[j] else: summary[text[0]][w]+=score[j] old_summary = summary summary = {} for key in old_summary: summary[key] = Counter(old_summary[key]).most_common(10) with open('output/locating/{}/{}.json'.format(self.args.dataset, self.args.word_score_name), 'w') as f: json.dump(summary, f, indent=4) def create_loader(self, file_name, shuffle): dataset = CLMDataset(file_name, self.tokenizer, self.args.device) loader = DataLoader(dataset=dataset, batch_size=self.args.batch_size, shuffle=shuffle, drop_last=False) return loader def generate_masked_utts(self): utt_intents = self.load_utt_intent(self.args.train) intent_related_words = self.load_semantic('output/locating/{}/{}.json'.format(self.args.dataset, self.args.word_score_name)) # print(intent_related_words) masked_utts = {'utt':[],'masked_word':[],'intent':[]} for utt_intent in tqdm(utt_intents): utt, intent = utt_intent if intent_related_words.get(intent, None) is None: continue for word in intent_related_words[intent]: if utt.count(word) == 1: masked_utts['utt'].append(utt.replace(word, '[MASK]')) masked_utts['masked_word'].append(word) masked_utts['intent'].append(intent) # for intent, num in intent_num.items(): # print(intent, num) df = pd.DataFrame(masked_utts, columns=list(masked_utts.keys()), index=None) df.to_csv('output/locating/{}/{}.csv'.format(self.args.dataset, self.args.masked_utt_name)) if __name__ == '__main__': l = LocatingModule() l.generate_word_score() l.generate_masked_utts()
11493266
import queue visited = [] MAX = 20 adj = {1: set([2,3,5]), 2: set([3]), 3: set([8,4]), 4:6} q = queue.Queue(MAX) def dfs(s): if s in visited : return visited.append(s) print(s) children = adj.get(s) if children != None: if isinstance(children,int): # Case : values are not a set dfs(children) else : # Case : values are a set for i in children: dfs(i) return def test(): global visited visited = [] print("dfs : ") dfs(1) test()
11493278
class Solution: def convert(self, s, numRows): if numRows == 1 or numRows >= len(s): return s row, direction, res = 0, -1, [""] * numRows for char in s: res[row] += char if row == 0 or row == numRows - 1: direction *= -1 row += direction return "".join(res)
11493288
import os import sys import argparse def genera(envF): """ Read genera per environment :param envF: :return: """ with open(envF, 'r') as f: if __name__ == '__main__': parser = argparse.ArgumentParser(description="Calculate shared genera per env") parser.add_argument('-f', help='file of environments and genera') args = parser.parse_args()
11493328
import numpy as np import multiprocessing import timeit def start(): a = np.random.rand(1000, 1000) b = np.random.rand(1000, 1000) np.multiply(a,b) start2 = timeit.timeit() for i in range(500): start() end = timeit.timeit() print(end-start2) start2 = timeit.timeit() pool = multiprocessing.Pool(multiprocessing.cpu_count()) liste = [pool.apply_async(start, ()) for i in range(500)] [p.get() for p in liste] end = timeit.timeit() print(end - start2)
11493338
import collections class Solution: def maxNumberOfBalloons(self, text: str) -> int: cnt = collections.Counter(text) ans = float('inf') for ch in 'ban': ans = min(ans, cnt.get(ch, 0)) for ch in 'lo': ans = min(ans, cnt.get(ch, 0) // 2) return ans
11493352
from numpy import zeros, copy from itertools import accumulate M = 1e9 + 7 class Solution: def numOfArrays(self, n: int, m: int, k: int) -> int: k, m = k + 1, m + 1 f1 = zeros((k, m)) f1[1, 1:] = 1 for _ in range(1, n): f2 = copy(f1) for j in range(1, k): item2s = list(accumulate(f1[j - 1])) for c in range(1, m): f2[j, c] = (f1[j, c] * c + item2s[c - 1]) % M f1 = f2 return int(sum(f1[k - 1, 1:]) % M)
11493378
import pandas as pd import numpy as np def dataset(): train_data = pd.read_csv('train.csv', index_col=None) direction_data = train_data['direction'] sentence_data = train_data['sentence'] rel_data = train_data['relation'] term1_data = train_data['term1'] term2_data = train_data['term2'] m = [] fd = pd.read_csv('train.csv', usecols = ['_unit_id']) count = 1 total = 0 ls = [] #print(len(sentence_data)) for i in fd['_unit_id']: ls.append(i) ls.append(0) # so that last set of sentences are counted j = 0 i = 0 for j in range(0, len(ls)-1): if ls[j] == ls[j+1]: count = count + 1 else: dir_data = direction_data.ix[i:i+count-1].sort_values() s = set(dir_data) a = [] for x in s: a.append(x) #print(len(a)) c1 = 0 c2 = 0 c3 = 0 if len(a) != 1: for j in dir_data: if a[0] == j: c1+=1 elif a[1] == j: c2+=1 else: c3+=1 else: c1 = count c2 = c3 = 0 if c1 > c2 and c1 > c3: m.append([a[0], sentence_data.ix[i], rel_data[i], term1_data[i], term2_data[i]]) elif c2 > c1 and c2 > c3: m.append([a[1], sentence_data.ix[i], rel_data[i], term1_data[i], term2_data[i]]) elif c3 > c1 and c3 > c2: m.append([a[2], sentence_data.ix[i], rel_data[i], term1_data[i], term2_data[i]]) i += count count = 1 columns= ['direction','sentence', 'relation', 'term1', 'term2'] data = pd.DataFrame(m, columns=columns) return data
11493410
from ..cw_model import CWModel class Report(CWModel): def __init__(self, json_dict=None): self.column_definitions = None # (JObject[]) self.row_values = None # (JObject[]) # initialize object with json dict super().__init__(json_dict)
11493416
import dataclasses import re from typing import Any, ClassVar, Match, Optional, Pattern, Set, Union BASE_PATTERN = re.compile( r"^arn:" r"(?P<partition>.+?):" r"(?P<service>.+?):" r"(?P<region>.*?):" r"(?P<account>.*?):" r"(?P<rest>.*)$" ) class InvalidArnException(Exception): """Raised when the value cannot be parsed as a valid ARN.""" def __init__(self, arn: str): self.arn = arn super().__init__(f"{arn} is not a valid ARN") class InvalidArnRestException(Exception): """Raised when the value can be parsed as a valid ARN but the rest cannot.""" def __init__(self, rest: str, class_name: str) -> None: self.rest = rest self.class_name = class_name super().__init__(f"{rest} is not a valid rest expression for type {class_name}") class ConflictingFieldNamesException(Exception): """A subclass tried to use reserved field names. The :py:attr:`partition`, :py:attr:`service`, :py:attr:`region`, and :py:attr:`account` names are reserved and cannot be used as attributes. """ def __init__(self, field_names: Set[str]) -> None: self.field_names = field_names super().__init__( f"Fields {', '.join(field_names)} are reserved and " f"cannot be used as field names" ) @dataclasses.dataclass() class Arn: """The base class that represents an AWS ARN. This class is meant to be used as a superclass for more specific ARN classes. Instantiating this class with a valid ARN will parse out the common fields (:py:attr:`partition`, :py:attr:`service`, :py:attr:`region`, and :py:attr:`account`), and will place the rest of the ARN string in the :py:attr:`rest` field. """ REST_PATTERN: ClassVar[Union[str, Pattern]] = re.compile(r"(?P<rest>.*)") """The pattern that parses the "rest" of the ARN. The "rest" of and ARN is the part that is specific to the AWS service that the ARN represents. When overriding in a subclass, this value can be either an `re.Pattern`_ or an ``str``. .. _re.Pattern: https://docs.python.org/3/library/re.html#regular-expression-objects """ input_arn: Any """The instance that was parsed, unchanged.""" partition: str = "" """The partition of the AWS of the resource.""" service: str = "" """The AWS service of the resource.""" region: str = "" """The AWS region in which the resource is located.""" account: str = "" # str because some pre-built resources have "aws" as the account """The AWS account ID of the resource.""" rest: str = dataclasses.field(init=False, default="") """The rest of the ARN, as matched by :py:const:`REST_PATTERN`.""" def __post_init__(self) -> None: if isinstance(self.input_arn, bytes): arn = self.input_arn.decode() elif isinstance(self.input_arn, str): arn = self.input_arn else: arn = str(self.input_arn) base_match = BASE_PATTERN.match(arn) if not base_match: raise InvalidArnException(arn) self._assign_fields_from_match(base_match) rest = base_match["rest"] rest_match = self.match_rest(rest) if not rest_match: raise InvalidArnRestException(rest, self.__class__.__name__) reserved_field_names = {f.name for f in dataclasses.fields(Arn) if f.init} subclass_field_names = set(rest_match.re.groupindex.keys()) conflicting_field_names = reserved_field_names & subclass_field_names if conflicting_field_names: raise ConflictingFieldNamesException(conflicting_field_names) self.assign_rest(rest_match) def __str__(self): return ( f"arn:" f"{self.partition}:" f"{self.service}:" f"{self.region}:" f"{self.account}:" f"{self.format_rest()}" ) def match_rest(self, rest: str) -> Optional[Match]: """Convert the rest of the ARN into an `re.Match`_. By default, matches the rest of the ARN against :py:const:`REST_PATTERN`. Override this metod to match against a pattern dynamically. For an example, see :py:meth:`arn.ecs.ServiceArn.match_rest`. .. _re.Match: https://docs.python.org/3/library/re.html#match-objects """ return re.match(self.REST_PATTERN, rest) def assign_rest(self, match: Match): """Assign an `re.Match`_'s groups to fields on ``self``. By default, assigns all named groups in :py:const:`REST_PATTERN` as strings. Override this method to cast group matches to a more appropriate type. For an example, see :py:meth:`arn.ecs.TaskDefinitionArn.assign_rest`. .. _re.Match: https://docs.python.org/3/library/re.html#match-objects """ self._assign_fields_from_match(match) def format_rest(self) -> str: """Produce a formatted representation of the rest of the ARN. This method is essentially the reverse of :py:meth:`match_rest` and :py:meth:`assign_rest`. By default returns :py:data:`rest`. Override this method to allow users to override specific fields and get a ``str(...)`` representation that includes the override. """ return self.rest def _assign_fields_from_match(self, match): for key in match.re.groupindex.keys(): if not getattr(self, key): setattr(self, key, match[key])
11493443
import datetime from dojo.tools.stackhawk.parser import StackHawkParser from dojo.models import Test, Finding from unittests.dojo_test_case import DojoTestCase class TestStackHawkParser(DojoTestCase): __test_datetime = datetime.datetime(2022, 2, 16, 23, 7, 19, 575000, datetime.timezone.utc) def test_invalid_json_format(self): testfile = open("unittests/scans/stackhawk/invalid.json") parser = StackHawkParser() with self.assertRaises(ValueError): parser.get_findings(testfile, Test()) def test_parser_ensures_data_is_for_stackhawk_before_parsing(self): testfile = open("unittests/scans/stackhawk/oddly_familiar_json_that_isnt_us.json") parser = StackHawkParser() with self.assertRaises(ValueError): parser.get_findings(testfile, Test()) def test_stackhawk_parser_with_no_vuln_has_no_findings(self): testfile = open("unittests/scans/stackhawk/stackhawk_zero_vul.json") parser = StackHawkParser() findings = parser.get_findings(testfile, Test()) testfile.close() self.assertEqual(0, len(findings)) def test_stackhawk_parser_with_one_high_vuln_has_one_findings(self): testfile = open("unittests/scans/stackhawk/stackhawk_one_vul.json") parser = StackHawkParser() findings = parser.get_findings(testfile, Test()) testfile.close() self.__assertAllEndpointsAreClean(findings) self.assertEqual(1, len(findings)) finding = findings[0] self.__assertFindingEquals( finding, "Anti CSRF Tokens Scanner", self.__test_datetime, "Secured Application", "Development", "High", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/20012", "20012", "10", False, False ) def test_stackhawk_parser_with_many_vuln_has_many_findings_and_removes_duplicates(self): testfile = open("unittests/scans/stackhawk/stackhawk_many_vul.json") parser = StackHawkParser() findings = parser.get_findings(testfile, Test()) testfile.close() self.__assertAllEndpointsAreClean(findings) self.assertEqual(6, len(findings)) self.__assertFindingEquals( findings[0], "Cookie Slack Detector", self.__test_datetime, "Secured Application", "Development", "Low", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/90027", "90027", "10", False, False ) self.__assertFindingEquals( findings[1], "Proxy Disclosure", self.__test_datetime, "Secured Application", "Development", "Medium", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/40025", "40025", "10", False, False ) self.__assertFindingEquals( findings[2], "Anti CSRF Tokens Scanner", self.__test_datetime, "Secured Application", "Development", "High", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/20012", "20012", "10", False, False ) self.__assertFindingEquals( findings[3], "Cross Site Scripting Weakness (Reflected in JSON Response)", self.__test_datetime, "Secured Application", "Development", "High", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/40012", "40012", "1", False, False ) self.__assertFindingEquals( findings[4], "Content Security Policy (CSP) Header Not Set", self.__test_datetime, "Secured Application", "Development", "Medium", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/10038", "10038", "12", False, False ) self.__assertFindingEquals( findings[5], "Permissions Policy Header Not Set", self.__test_datetime, "Secured Application", "Development", "Low", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/10063", "10063", "12", False, False ) def test_that_a_scan_import_updates_the_test_description(self): testfile = open("unittests/scans/stackhawk/stackhawk_zero_vul.json") parser = StackHawkParser() test = Test() parser.get_findings(testfile, test) testfile.close() self.assertEqual( test.description, 'View scan details here: ' + '[https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27]' + '(https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27)' ) def test_that_a_scan_with_all_false_positive_endpoints_on_a_finding_marks_as_false_positive(self): testfile = open("unittests/scans/stackhawk/stackhawk_one_vuln_all_endpoints_false_positive.json") parser = StackHawkParser() findings = parser.get_findings(testfile, Test()) testfile.close() self.__assertAllEndpointsAreClean(findings) self.assertEqual(1, len(findings)) self.__assertFindingEquals( findings[0], "Cookie Slack Detector", self.__test_datetime, "Secured Application", "Development", "Low", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/90027", "90027", "3", True, False ) def test_that_a_scan_with_all_risk_accepted_endpoints_on_a_finding_marks_as_risk_accepted(self): testfile = open("unittests/scans/stackhawk/stackhawk_one_vuln_all_endpoints_risk_accepted.json") parser = StackHawkParser() findings = parser.get_findings(testfile, Test()) testfile.close() self.__assertAllEndpointsAreClean(findings) self.assertEqual(1, len(findings)) self.__assertFindingEquals( findings[0], "Cookie Slack Detector", self.__test_datetime, "Secured Application", "Development", "Low", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/90027", "90027", "3", False, True ) def test_that_a_scan_with_endpoints_in_differing_statuses_does_not_mark_as_risk_accepted_or_false_positive(self): testfile = open("unittests/scans/stackhawk/stackhawk_one_vuln_all_endpoints_have_different_status.json") parser = StackHawkParser() findings = parser.get_findings(testfile, Test()) testfile.close() self.__assertAllEndpointsAreClean(findings) self.assertEqual(1, len(findings)) self.__assertFindingEquals( findings[0], "Cookie Slack Detector", self.__test_datetime, "Secured Application", "Development", "Low", "https://app.stackhawk.com/scans/e2ff5651-7eef-47e9-b743-0c2f7d861e27/finding/90027", "90027", "3", False, False ) def __assertFindingEquals( self, actual_finding: Finding, title, date: datetime.datetime, application_name, environment, severity, finding_url, finding_id, count, false_positive, risk_accepted ): self.assertEqual(title, actual_finding.title) self.assertEqual(date, actual_finding.date) self.assertEqual(application_name, actual_finding.component_name) self.assertEqual(environment, actual_finding.component_version) self.assertEqual(severity, actual_finding.severity) self.assertEqual("View this finding in the StackHawk platform at:\n[" + finding_url + '](' + finding_url + ')', actual_finding.description) self.assertRegexpMatches( actual_finding.steps_to_reproduce, "Use a specific message link and click 'Validate' to see the cURL!.*" ) self.assertFalse(actual_finding.static_finding) self.assertTrue(actual_finding.dynamic_finding) self.assertEqual(finding_id, actual_finding.vuln_id_from_tool) self.assertEqual(count, actual_finding.nb_occurences) self.assertEqual(application_name, actual_finding.service) self.assertEqual(false_positive, actual_finding.false_p) self.assertEqual(risk_accepted, actual_finding.risk_accepted) # The following fields should be not be set from this parser. self.assertIsNone(actual_finding.unique_id_from_tool) def __assertAllEndpointsAreClean(self, findings): for finding in findings: for endpoint in finding.unsaved_endpoints: endpoint.clean()
11493478
import argparse import logging import os import subprocess from concurrent import futures from enum import Enum from os import PathLike from pathlib import Path from tempfile import NamedTemporaryFile from threading import Thread from typing import List from buglab.controllers.helper.randombugselectorserver import random_bug_selector_server from buglab.data.deduplication import DuplicationIndex from buglab.utils.logging import configure_logging LOGGER = logging.getLogger(__name__) class ExtractJob(Enum): BUG = 1 TYPE = 2 def create_container_and_extract( package_name: str, target_dir: PathLike, bug_selector_server_address: str, extract_job: ExtractJob ): docker_command = f'docker run --network="host" --rm -it -v {target_dir}:/data/targetDir buglab-base:latest ' if extract_job == ExtractJob.BUG: docker_command += ( f"python3.8 -m buglab.controllers.packageextracttodisk {package_name} " f"--bug-selector-server {bug_selector_server_address}" ) elif extract_job == ExtractJob.TYPE: docker_command += f"python3.8 -m buglab.controllers.typelessextracttodisk {package_name} " else: raise ValueError(f"Unknown extraction job type: {extract_job}") _ = subprocess.run(docker_command, shell=True) def extract_from_packages( packages: List[str], target_dir: PathLike, bug_selector_server_address: str, extract_job: ExtractJob ): with futures.ThreadPoolExecutor(max_workers=os.cpu_count()) as executor: all_futures = { executor.submit( create_container_and_extract, pkg, target_dir, bug_selector_server_address, extract_job ): pkg for pkg in packages } for future in futures.as_completed(all_futures): try: _ = future.result() except Exception as exc: LOGGER.exception("Failure for `%s` package", all_futures[future], exc_info=exc) if __name__ == "__main__": configure_logging() parser = argparse.ArgumentParser(description="Orchestrator to extract graphs across multiple packages.") parser.add_argument( "package_list_path", type=str, help="the path to a txt file containing the names of the packages to be considered", ) parser.add_argument("target_dir", type=str, help="the target directory to store the results.") parser.add_argument( "--extract-types", action="store_true", help="Set this flag if you want to remove type information instead of extracting bugs.", ) args = parser.parse_args() f = NamedTemporaryFile() duplication_index = DuplicationIndex(Path(f.name)) duplication_server_thread = Thread(target=lambda: duplication_index.server(address="tcp://*:5555"), daemon=True) duplication_server_thread.start() rewrite_selector_server_port: str = "8345" extract_job = ExtractJob.TYPE if not args.extract_types: extract_job = ExtractJob.BUG server_thread = Thread( target=lambda: random_bug_selector_server("tcp://*:" + rewrite_selector_server_port), daemon=True ) server_thread.start() all_packages = [] with open(args.package_list_path) as f: for line in f.readlines(): pkg_name = line.strip() if len(pkg_name) > 0: all_packages.append(pkg_name) extract_from_packages(all_packages, args.target_dir, "tcp://localhost:" + rewrite_selector_server_port, extract_job)
11493505
import re import pandas as pd from config_db import DB_CONNECTION __all__ = [] OPTIONS = dict( if_exists='replace', index=False ) def gsheet_csv_url(url): """ Returns the url for the Google Sheet csv export. Parameters ---------- url : str The editor url string as found when viewing the sheet in a browser. """ def get_sheet(): for i, x in enumerate(s): if x == 'gid': return s[i+1] raise ValueError('Sheet ID not found in url {}'.format(url)) s = re.split('/|#|=|&', url) key = s[5] sheet = get_sheet() return 'https://docs.google.com/spreadsheets/d/{}/export?gid={}&format=csv'.format(key, sheet) def make_unique_columns(df): """ Renames the data frame columns to be unique when exported to an SQL database. Parameters ---------- df : :class:`pandas.DataFrame` The data frame. """ names = {} lnames = set() for x in list(df): n = x # Get a unique name while n.lower() in lnames: n = n + '_' lnames.add(n.lower()) names[x] = n df.rename(columns=names, inplace=True) return df def rename_columns_15_0(df): """ Renames the version 15.0 data frame columns. Parameters ---------- df : :class:`pandas.DataFrame` The data frame. """ names = { 'AISC_Manual_Label': 'name', 'W': 'unit_weight', 'A': 'area', 'Ix': 'inertia_x', 'Iy': 'inertia_y', 'Iz': 'inertia_z', 'Zx': 'plast_sect_mod_x', 'Zy': 'plast_sect_mod_y', 'Sx': 'elast_sect_mod_x', 'Sy': 'elast_sect_mod_y', 'Sz': 'elast_sect_mod_z', 'rx': 'gyradius_x', 'ry': 'gyradius_y', 'rz': 'gyradius_z', 'J': 'inertia_t', } df.rename(columns=names, inplace=True) return df def write_aisc_metric_15_0(): """ Writes the 'aisc_metric_15_0' table to the database. """ url = 'https://docs.google.com/spreadsheets/d/1RwpcQxKsQmb_ylxR5Zx4JYWf9_A4Cd-6KrUrXt3ynls/edit#gid=55672305' table = 'aisc_metric_15_0' url = gsheet_csv_url(url) df = pd.read_csv(url) rename_columns_15_0(df) make_unique_columns(df) df.to_sql(table, DB_CONNECTION, **OPTIONS) def write_aisc_imperial_15_0(): """ Writes the 'aisc_imperial_15_0' table to the database. """ url = 'https://docs.google.com/spreadsheets/d/1RwpcQxKsQmb_ylxR5Zx4JYWf9_A4Cd-6KrUrXt3ynls/edit#gid=1797343786' table = 'aisc_imperial_15_0' url = gsheet_csv_url(url) df = pd.read_csv(url) rename_columns_15_0(df) make_unique_columns(df) df.to_sql(table, DB_CONNECTION, **OPTIONS) def write_database(): """ Writes all data to the database. """ write_aisc_metric_15_0() write_aisc_imperial_15_0() if __name__ == '__main__': write_database()
11493509
from django.urls import path from custom_auth.oauth.views import login_with_github, login_with_google app_name = "oauth" urlpatterns = [ path("google/", login_with_google, name="google-oauth-login"), path("github/", login_with_github, name="github-oauth-login"), ]
11493574
import random class Position: def __init__(self, x, y): self.x = x self.y = y def __hash__(self): return hash((self.x, self.y)) def __eq__(self, other): return self.x == other.x and self.y == other.y def __repr__(self): return "Position=[x={},y={}]".format(self.x, self.y) def get_potential_knight_moves(start, size, visited): moves = list() moves.append(Position(start.x + 1, start.y + 2)) moves.append(Position(start.x + 1, start.y - 2)) moves.append(Position(start.x - 1, start.y + 2)) moves.append(Position(start.x - 1, start.y - 2)) moves.append(Position(start.x + 2, start.y + 1)) moves.append(Position(start.x + 2, start.y - 1)) moves.append(Position(start.x - 2, start.y + 1)) moves.append(Position(start.x - 2, start.y - 1)) valid_moves = [pos for pos in moves if pos.x >= 0 and pos.x < size and pos.y >= 0 and pos.y < size and pos not in visited] return valid_moves def run_knights_tour(start, size, visited): if len(visited) == size * size: return 1 moves = get_potential_knight_moves(start, size, visited) count = 0 for move in moves: tmp_visted = visited.copy() tmp_visted.add(move) count += run_knights_tour(move, size, tmp_visted) return count def count_knights_tours(size): count = 0 for i in range(size): for j in range(size): start = Position(i, j) count += run_knights_tour(start, size, set([start])) return count assert count_knights_tours(1) == 1 assert count_knights_tours(2) == 0 assert count_knights_tours(3) == 0 assert count_knights_tours(4) == 0 assert count_knights_tours(5) == 1728
11493590
from .base import * # SECURITY WARNING: don't run with debug turned on in production! DEBUG = True ALLOWED_HOSTS = ['*'] DATABASES = { 'default': { 'ENGINE': 'django.db.backends.sqlite3', 'NAME': 'db.db', } } # Required to serve the files while we use Gunicorn in dev env MIDDLEWARE += [ 'whitenoise.middleware.WhiteNoiseMiddleware', ]
11493634
import pytest from dbnd._core.settings.tracking_config import ( TrackingConfig, ValueTrackingLevel, get_value_meta, ) from targets import target as target_factory from targets.value_meta import ValueMetaConf from targets.values import ListValueType, ObjectValueType, StrValueType, TargetValueType class TestTrackingConfig: @pytest.mark.parametrize( "value, value_type, target, expected_value_preview, expected_data_schema", [ (10, None, None, "10", {"type": "int"}), (10, ObjectValueType(), None, "10", {"type": "int"}), (10, TargetValueType(), target_factory("/path"), "10", {"type": "int"}), (10, StrValueType(), None, "10", {"type": "str"}), ([10], ListValueType(), None, "[10]", {"type": "List"}), ], ) def test_get_value_meta( self, value, value_type, target, expected_value_preview, expected_data_schema, ): tracking_config = TrackingConfig.current() tracking_config.value_reporting_strategy = ValueTrackingLevel.ALL result = get_value_meta( value, ValueMetaConf(), tracking_config, value_type=value_type, target=target, ) assert result.value_preview == expected_value_preview assert result.data_schema == expected_data_schema
11493637
from ..utility.expr_wrap_util import symbolic from ..expr import BVV, BVS from ..utility.models_util import get_arg_k from ..sym_state import State MAX_READ = 100 def read_handler(state: State, view): fd = get_arg_k(state, 1, 4, view) buf = get_arg_k(state, 2, state.arch.bits() // 8, view) count = get_arg_k(state, 3, 4, view) assert not symbolic(fd) or not state.solver.symbolic(fd) fd = fd.value assert state.os.is_open(fd) if symbolic(count): count = state.solver.max(count) count = MAX_READ if count > MAX_READ else count else: count = count.value res = state.os.read(fd, count) for i, b in enumerate(res): state.mem.store(buf + i, b) state.events.append( "read from fd %d, count %d" % (fd, count) ) return BVV(count, 32) def write_handler(state: State, view): fd = get_arg_k(state, 1, 4, view) buf = get_arg_k(state, 2, state.arch.bits() // 8, view) count = get_arg_k(state, 3, 4, view) assert not symbolic(fd) or not state.solver.symbolic(fd) fd = fd.value if symbolic(count): count = state.solver.max(count) count = MAX_READ if count > MAX_READ else count else: count = count.value data = [] for i in range(count): b = state.mem.load(buf + i, 1) data.append(b) state.os.write(fd, data) state.events.append( "write to fd %d, count %d" % (fd, count) ) return BVV(count, 32) stat_idx = 0 def _stat(state: State, statbuf): global stat_idx long_t = state.arch.bits() int_t = 32 st_dev = BVS('stat_st_dev_%d' % stat_idx, long_t) st_ino = BVS('stat_st_ino_%d' % stat_idx, long_t) st_mode = BVS('stat_st_mode_%d' % stat_idx, long_t) st_nlink = BVS('stat_st_nlink_%d' % stat_idx, long_t) st_uid = BVS('stat_st_uid_%d' % stat_idx, int_t) st_gid = BVS('stat_st_gid_%d' % stat_idx, int_t) st_rdev = BVS('stat_st_rdev_%d' % stat_idx, long_t) st_size = BVS('stat_st_size_%d' % stat_idx, long_t) st_blksize = BVS('stat_st_blksize_%d' % stat_idx, long_t) st_blocks = BVS('stat_st_blocks_%d' % stat_idx, long_t) st_atim_tv_sec = BVS('stat_atim.sec_%d' % stat_idx, long_t) st_atim_tv_nsec = BVS('stat_atim.nsec_%d' % stat_idx, long_t) st_mtim_tv_sec = BVS('stat_mtim.sec_%d' % stat_idx, long_t) st_mtim_tv_nsec = BVS('stat_mtim.nsec_%d' % stat_idx, long_t) st_ctim_tv_sec = BVS('stat_ctim.sec_%d' % stat_idx, long_t) st_ctim_tv_nsec = BVS('stat_ctim.nsec_%d' % stat_idx, long_t) stat_idx += 1 state.mem.store(statbuf + 0, st_dev, state.arch.endness()) state.mem.store(statbuf + 8, st_ino, state.arch.endness()) state.mem.store(statbuf + 16, st_nlink, state.arch.endness()) state.mem.store(statbuf + 24, st_mode, state.arch.endness()) state.mem.store(statbuf + 32, st_uid, state.arch.endness()) state.mem.store(statbuf + 36, st_gid, state.arch.endness()) state.mem.store(statbuf + 40, BVV(0, 8*8)) # padding state.mem.store(statbuf + 48, st_rdev, state.arch.endness()) state.mem.store(statbuf + 56, st_size, state.arch.endness()) state.mem.store(statbuf + 64, st_blksize, state.arch.endness()) state.mem.store(statbuf + 72, st_blocks, state.arch.endness()) state.mem.store(statbuf + 80, st_atim_tv_sec, state.arch.endness()) state.mem.store(statbuf + 88, st_atim_tv_nsec, state.arch.endness()) state.mem.store(statbuf + 96, st_mtim_tv_sec, state.arch.endness()) state.mem.store(statbuf + 104, st_mtim_tv_nsec, state.arch.endness()) state.mem.store(statbuf + 112, st_ctim_tv_sec, state.arch.endness()) state.mem.store(statbuf + 120, st_ctim_tv_nsec, state.arch.endness()) state.mem.store(statbuf + 128, BVV(0, 8*16)) # reserved (zero (?)) return BVV(0, 32) def stat_handler(state: State, view): global stat_idx pathname = get_arg_k(state, 1, state.arch.bits() // 8, view) statbuf = get_arg_k(state, 2, state.arch.bits() // 8, view) path = "" if not symbolic(pathname): i = 0 c = state.mem.load(pathname, 1) while not symbolic(c) and c.value != 0 and i < 100: path += chr(c.value) i += 1 c = state.mem.load(pathname+i, 1) else: path = "<symbolic>" state.events.append( "stat on %s" % path ) return _stat(state, statbuf) def xstat_handler(state: State, view): version = get_arg_k(state, 1, 4, view) pathname = get_arg_k(state, 2, state.arch.bits() // 8, view) statbuf = get_arg_k(state, 3, state.arch.bits() // 8, view) path = "" if not symbolic(pathname): i = 0 c = state.mem.load(pathname, 1) while not symbolic(c) and c.value != 0 and i < 100: path += chr(c.value) i += 1 c = state.mem.load(pathname+i, 1) else: path = "<symbolic>" if not symbolic(version): version = str(version.value) else: version = "<symbolic>" state.events.append( "__xstat on %s. version %s" % (path, version) ) return _stat(state, statbuf)
11493750
from __future__ import absolute_import from __future__ import division from __future__ import print_function import numpy as np from six.moves import xrange # TODO # - Proper tests of everything # - be clearer about meaning of t_elapsed, t_ix and either (t) # - Time Since Event is a ticking bomb. Needs better naming/definitions # to ensure that it's either inverse TTE or a feature or if they coincide. def roll_fun(x, size, fun=np.mean, reverse=False): """Like cumsum but with any function `fun`. """ y = np.copy(x) n = len(x) size = min(size, n) if size <= 1: return x for i in xrange(size): y[i] = fun(x[0:(i + 1)]) for i in xrange(size, n): y[i] = fun(x[(i - size + 1):(i + 1)]) return y def carry_forward_if(x, is_true): """Locomote forward `x[i]` if `is_true[i]`. remain x untouched before first pos of truth. :param Array x: object whos elements are to carry forward :param Array is_true: same length as x containing true/false boolean. :return Array x: forwarded object """ for i in xrange(len(x)): if is_true[i]: cargo = x[i] if cargo is not None: x[i] = cargo return x def carry_backward_if(x, is_true): """Locomote backward `x[i]` if `is_true[i]`. remain x untouched after last pos of truth. :param Array x: object whos elements are to carry backward :param Array is_true: same length as x containing true/false boolean. :return Array x: backwarded object """ for i in xrange(reversed(len(x))): if is_true[i]: cargo = x[i] if cargo is not None: x[i] = cargo return x def steps_since_true_minimal(is_event): """(Time) since event over discrete (padded) event vector. :param Array is_event: a vector of 0/1s or boolean :return Array x: steps since is_event was true """ n = len(is_event) z = -1 # at the latest on step before x = np.int32(is_event) for i in xrange(n): if is_event[i]: z = i x[i] = i - z return x def steps_to_true_minimal(is_event): """(Time) to event for discrete (padded) event vector. :param Array is_event: a vector of 0/1s or boolean :return Array x: steps until is_event is true """ n = len(is_event) z = n # at the earliest on step after x = np.int32(is_event) for i in reversed(xrange(n)): if is_event[i]: z = i x[i] = z - i return x def get_tte_discrete(is_event, t_elapsed=None): """Calculates discretely measured tte over a vector. :param Array is_event: Boolean array :param IntArray t_elapsed: integer array with same length as `is_event`. If none, it will use `xrange(len(is_event))` :return Array tte: Time-to-event array (discrete version) - Caveats tte[i] = numb. timesteps to timestep with event Step of event has tte = 0 \ (event happened at time [t,t+1)) tte[-1]=1 if no event (censored data) """ n = len(is_event) tte = np.int32(is_event) stepsize = 1 if t_elapsed is None: t_elapsed = xrange(n) t_next = t_elapsed[-1] + stepsize for i in reversed(xrange(n)): if is_event[i]: t_next = t_elapsed[i] tte[i] = t_next - t_elapsed[i] return tte def get_tte_continuous(is_event, t_elapsed): """Calculates time to (pointwise measured) next event over a vector. :param Array is_event: Boolean array :param IntArray t_elapsed: integer array with same length as `is_event` that supports vectorized subtraction. If none, it will use `xrange(len(is_event))` :return Array tte: Time-to-event (continuous version) TODO:: Should support discretely sampled, continuously measured TTE .. Caveats:: tte[i] = time to *next* event at time t[i] (t[i] is exactly at event&/or query) tte[-1]=0 always (since last time is a *point*) Last datpoints are right censored. """ n = len(is_event) if t_elapsed is None: t_elapsed = np.int32(xrange(n)) t_next = t_elapsed[-1] # lazy initialization to autoinit if difftime tte = t_elapsed - t_next for i in reversed(xrange(n)): tte[i] = t_next - t_elapsed[i] if is_event[i]: t_next = t_elapsed[i] return tte def get_tte(is_event, discrete_time, t_elapsed=None): """ wrapper to calculate *Time To Event* for input vector. :param Boolean discrete_time: if `True`, use `get_tte_discrete`. If `False`, use `get_tte_continuous`. """ if discrete_time: return get_tte_discrete(is_event, t_elapsed) else: return get_tte_continuous(is_event, t_elapsed) def get_tse(is_event, t_elapsed=None): """ Wrapper to calculate *Time Since Event* for input vector. Inverse of tte. Safe to use as a feature. Always "continuous" method of calculating it. tse >0 at time of event (if discrete we dont know about the event yet, if continuous we know at record of event so superfluous to have tse=0) tse = 0 at first step :param Array is_event: Boolean array :param IntArray t_elapsed: None or integer array with same length as `is_event`. * If none, it will use `t_elapsed.max() - t_elapsed[::-1]`. .. TODO:: reverse-indexing is pretty slow and ugly and not a helpful template for implementing in other languages. """ if t_elapsed is not None: t_elapsed = t_elapsed.max() - t_elapsed[::-1] return get_tte_continuous(is_event[::-1], t_elapsed)[::-1] def get_is_not_censored(is_event, discrete_time=True): """ Calculates non-censoring indicator `u` for one vector. :param array is_event: logical or numeric array indicating event. :param Boolean discrete_time: if `True`, last observation is conditionally censored. """ n = len(is_event) is_not_censored = np.copy(is_event) if discrete_time: # Last obs is conditionally censored event_seen = is_event[-1] for i in reversed(xrange(n)): if is_event[i] and not event_seen: event_seen = is_event[i] is_not_censored[i] = event_seen else: # Last obs is always censored event_seen = False for i in reversed(xrange(n)): is_not_censored[i] = event_seen if is_event[i] and not event_seen: event_seen = is_event[i] return is_not_censored
11493754
from typing import Optional from dissect.cstruct import Instance from structlog import get_logger from unblob.extractors.command import Command from unblob.file_utils import InvalidInputFormat from ...models import File, HexString, StructHandler, ValidChunk logger = get_logger() VALID_NT_SIGNATURES = [0x28, 0x29] VALID_MEDIAS = [0xF0, 0xF8, 0xF9, 0xFA, 0xFB, 0xFC, 0xFD, 0xFE, 0xFF] class FATHandler(StructHandler): NAME = "fat" PATTERNS = [ HexString( """ // An initial x86 short jump instruction // OEMName (8 bytes) // BytesPerSec (2 bytes) // SecPerClus (1 byte) "Must be one of 1, 2, 4, 8, 16, 32, 64, 128." // 495 (0x1EF) bytes of whatever // 55 AA is the "signature". "This will be the end of the sector only in case the // sector size is 512." ( EB | E9 ) [13] ( 01 | 02 | 04 | 08 | 10 | 20 | 40 | 80 ) [495] 55 AA """ ) ] C_DEFINITIONS = r""" // Common between FAT12, FAT16 and FAT32. typedef struct bios_param_common { char JmpBoot[3]; // An x86 JMP - e.g. EB 3C 90 ("One finds either eb xx 90, or e9 xx xx.") char OEMName[8]; // OEM name/version (E.g. "IBM 3.3", "IBM 20.0", "MSDOS5.0", "MSWIN4.0".) // "BIOS Parameter Block" starts here uint16 BytesPerSec; // Almost always 512? Microsoft operating systems // will properly support 1024, 2048, and 4096. uint8 SecPerClus; // The legal values are 1, 2, 4, 8, 16, 32, 64, // and 128. uint16 RsvdSecCnt; // Must not be 0. Should be 1 on FAT12/16? 32 on FAT32? uint8 NumFATs; // Should always be 2? uint16 RootEntCnt; // Number of 32-byte dir entries. Must be 0 for FAT32. uint16 TotSectors; // Old 16-bit count of all sectors on volume // This field can be 0; if it is 0, then TotSec32 must be // non-zero. For FAT32, this field must be 0. For FAT12 and // FAT16 volumes, this field contains the sector count, and // TotSec32 is 0 if the total sector count fits (is less than // 0x10000). uint8 Media; // The valid values for this field is 0xF0, 0xF8, 0xF9, 0xFA, 0xFB, // 0xFC, 0xFD, 0xFE and 0xFF. uint16 FATSz16; // This field is the FAT12/FAT16 16-bit count of sectors occupied by // ONE FAT. On FAT32 volumes this field must be 0, and // FATSz32 contains the FAT size count. uint16 SecPerTrk; // Sectors per track for interrupt 0x13. uint16 NumHeads; // Number of heads for interrupt 0x13 } bios_param_common_t; // BIOS params for FAT16. typedef struct fat12_16_bootsec { bios_param_common_t common; uint32 NumHidden; uint32 NumSectors; uint8 DrvNum; uint8 Reserved1; uint8 BootSig; // If it's 0x29 (or 0x28 on NT), means that the next 3 fields are present char VolID[4]; char VolLab[11]; char FileSysType[8]; // Filesystem type (E.g. "FAT12 ", "FAT16 ", "FAT ", or all zero.) } fat12_16_bootsec_t; // BIOS params for FAT32. typedef struct fat32_bootsec { bios_param_common_t common; uint32 Num_Hidden; uint32 TotSec32; uint32 FATSz32; uint16 ExtFlags; uint16 FSVer; uint32 RootClus; uint16 FSInfo; uint16 BkBootSec; uint8 Reserved[12]; uint8 DrvNum; uint8 Reserved1; uint8 BootSig; char VolID[4]; char VolLab[11]; char FileSysType[8]; } fat32_bootsec_t; typedef struct fat_unknown { bios_param_common_t common; } fat_unknown_t; """ EXTRACTOR = Command("7z", "x", "-y", "{inpath}", "-o{outdir}") def valid_name(self, name: bytes) -> bool: try: name.decode("utf-8") except UnicodeDecodeError: return False return True def valid_fat32_header(self, header: Instance) -> bool: if header.common.RootEntCnt != 0: return False if header.common.TotSectors != 0: return False if header.common.FATSz16 != 0: return False return True def valid_header(self, header: Instance) -> bool: if not self.valid_name(header.common.OEMName): return False if header.common.BytesPerSec % 2 != 0: return False if header.common.RsvdSecCnt == 0: return False if header.common.Media not in VALID_MEDIAS: return False if header.BootSig not in VALID_NT_SIGNATURES: return False if header.FileSysType not in (b"FAT12 ", b"FAT16 "): return self.valid_fat32_header(header) return True def calculate_chunk(self, file: File, start_offset: int) -> Optional[ValidChunk]: header = self.cparser_le.fat12_16_bootsec_t(file) if header.FileSysType in (b"FAT12 ", b"FAT16 "): if header.common.TotSectors == 0: sector_count = header.NumSectors else: sector_count = header.common.TotSectors else: file.seek(start_offset) header = self.cparser_le.fat32_bootsec_t(file) sector_count = header.TotSec32 if not self.valid_header(header): raise InvalidInputFormat("Invalid FAT header.") logger.debug("FAT header parsed", header=header, _verbosity=3) if sector_count == 0x0: sector_count = header.common.TotSectors size = header.common.BytesPerSec * sector_count return ValidChunk( start_offset=start_offset, end_offset=start_offset + size, )
11493758
from deepmatch.models import NCF from ..utils import get_xy_fd_ncf def test_NCF(): model_name = "NCF" x, y, user_feature_columns, item_feature_columns = get_xy_fd_ncf(False) model = NCF(user_feature_columns, item_feature_columns, ) model.compile('adam', "binary_crossentropy") model.fit(x, y, batch_size=10, epochs=2, validation_split=0.5) if __name__ == "__main__": pass