markdown stringlengths 0 1.02M | code stringlengths 0 832k | output stringlengths 0 1.02M | license stringlengths 3 36 | path stringlengths 6 265 | repo_name stringlengths 6 127 |
|---|---|---|---|---|---|
Lecture de la vรฉritable rรฉponse | AUS_response = pd.read_csv('AUS_test_Rain_tomorrow.csv', index_col=['Date'], parse_dates=True)
y_true = (AUS_response['RainTomorrow'] == 'Yes').astype(int) | _____no_output_____ | MIT | projet2/projet2-solution.ipynb | tiombo/Projet-Final |
Calcul du log-loss | log_loss(y_true, y_test_pred_proba_best) | _____no_output_____ | MIT | projet2/projet2-solution.ipynb | tiombo/Projet-Final |
Aire sous la courbe | print(f'AUC = {roc_auc_score(y_true, y_test_pred_proba_best)}') | AUC = 0.8737077575165082
| MIT | projet2/projet2-solution.ipynb | tiombo/Projet-Final |
Courbe ROC | fpr_rf, tpr_rf, thresholds = roc_curve(y_true, y_test_pred_proba_best)
fig = plt.figure(4, figsize=(6, 6))
plt.plot([0, 1], [0, 1], 'k--')
plt.plot(fpr_rf, tpr_rf, label='Meilleur modรจle')
plt.xlabel('False positive rate')
plt.ylabel('True positive rate')
plt.title('ROC curve')
plt.show() | _____no_output_____ | MIT | projet2/projet2-solution.ipynb | tiombo/Projet-Final |
Score F1 | f1_score(y_true, y_test_pred_proba_best > selected_threshold) # Attention, utiliser le seuil trouvรฉ par validation croisรฉe ! | _____no_output_____ | MIT | projet2/projet2-solution.ipynb | tiombo/Projet-Final |
Matrice de confusion | fig = plt.figure(3, figsize=(6, 6))
cnf_matrix = confusion_matrix(y_true, y_test_pred_proba_best > selected_threshold)
np.set_printoptions(precision=2)
plot_confusion_matrix(cnf_matrix, classes=['0','1'], title='Matrice de confusion')
# Accuracy
(26094+7500)/(26094+7500+7045+2018) | _____no_output_____ | MIT | projet2/projet2-solution.ipynb | tiombo/Projet-Final |
**Data Pre Processing** | DATA_DIR = "Beijing-Pollution-DataSet/"
from pandas import read_csv
from datetime import datetime
from random import randint
def select_month(sequences, n_samples=250):
X, y = list(), list()
rand_hour = randint(0, 24)
rand_day = randint(0, 7)
for i in range(0, n_samples):
start_ix = rand_hour +... | _____no_output_____ | MIT | Q1 PartD/Q1 PartD - Monthly/MiniProj_RNN_ADAgrad_MSE_Q1_PartD_Pytorch_Monthly.ipynb | parhamzm/Beijing-Weather-Prediction |
Preprocessing | raw_data.describe(include='all')
data = raw_data.drop(['Model'],axis=1)
data.describe(include='all')
# dropped the Model category because it is irrelevant to increasing the accuracy of my model.
data.isnull().sum()
#this shows the missing values in my dataset
data_no_mv = data.dropna(axis=0)
#I drop the missing values ... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
PDFs Here I check the Probability Distribution Functions (PDF) of the Independent Variables Price, Year, Mileage, and Engine Volume to identify and weed out the Outliers. They can adversely affect the accuracy of my Regression model because a Regression attempts to draw a line closest to all the data; including the Ou... | sns.distplot(data_no_mv['Price'])
q = data_no_mv['Price'].quantile(0.99)
data_1 = data_no_mv[data_no_mv['Price']<q]
data_1.describe(include='all')
# I decided to exclude the observations in the 99th percentile and above to get rid of the Outliers.
sns.distplot(data_1['Price'])
# Now the Price variable only includes obs... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
Checking the OLS assumptions Distribution | f, (ax1, ax2, ax3) = plt.subplots(1, 3, sharey=True, figsize =(15,3))
ax1.scatter(data_cleaned['Year'],data_cleaned['Price'])
ax1.set_title('Price and Year')
ax2.scatter(data_cleaned['EngineV'],data_cleaned['Price'])
ax2.set_title('Price and EngineV')
ax3.scatter(data_cleaned['Mileage'],data_cleaned['Price'])
ax3.set_t... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
Multicollinearity | data_cleaned.columns.values
from statsmodels.stats.outliers_influence import variance_inflation_factor
variables = data_cleaned[['Mileage','Year','EngineV']]
vif = pd.DataFrame()
vif["VIF"] = [variance_inflation_factor(variables.values, i) for i in range(variables.shape[1])]
vif["features"] = variables.columns
# Throug... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
Training my Regression Model | targets = data_preprocessed['log_price']
inputs = data_preprocessed.drop(['log_price'], axis=1)
# I removed log_price in the inputs to exclude the transformed dependent variable from my inputs.
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler ()
scaler.fit(inputs)
inputs_scaled = scaler.transfor... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
Using Residuals to check the Model | sns.distplot(y_train-y_hat)
plt.title("Residuals PDF", size=20)
#to check whether the Residuals is normally distributed and the variability of the outcome
reg.score(x_train, y_train)
reg.intercept_
# The intercept or bias calibrates the model: without it, each feature will be off the mark.
reg.coef_
reg_summary=pd.Data... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
Testing my Model | y_hat_test = reg.predict(x_test)
plt.scatter(y_test, y_hat_test, alpha=0.2)
plt.xlabel('Targets(y_test)', size=20)
plt.ylabel('Predictions(y_hat_test)', size=20)
plt.xlim(6, 13)
plt.ylim(6, 13)
plt.show()
df_pf = pd.DataFrame(np.exp(y_hat_test), columns=['Prediction'])
#this returns the exponential of Y Hat Test and re... | _____no_output_____ | Apache-2.0 | Linear Regression w sklearn Case Exercise.ipynb | jovemmanuelre/Practical-Case-Example-Regression-with-sklearn |
Simple Sentiment Analysis PipelineHere we train simple 2 layer neural network for sentiment analysis. - Model: 2 Fully Connected layer NN (PyTorch)- Dataset: Sentiment Analysis- Embedding: spacy en_core_web_lg (mean aggregated embeddings of the text) Install Requirements from [repository](https://github.com/AnzorGoza... | !wget https://raw.githubusercontent.com/AnzorGozalishvili/active_learning_playground/main/requirements.txt
!pip install -r requirements.txt
!rm requirements.txt
!pip install spacy-sentence-bert==0.1.2 | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Imports | # system
import os
import sys
# data and models
import numpy as np
import pandas as pd
import scipy
# utilities
import random
import re
import datetime
# text embeddings
import spacy
import spacy_sentence_bert
# scikit-learn stuff
import sklearn
from sklearn.model_selection import train_test_split
from sklearn.metr... | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Set Random SeedsFor reproducibility we set several random seeds which are recommended by PyTorch. ([See here](https://pytorch.org/docs/stable/notes/randomness.html)) | random.seed(hash("setting random seeds") % 2**32 - 1)
np.random.seed(hash("improves reproducibility") % 2**32 - 1)
torch.manual_seed(hash("PyTorch") % 2**32 - 1)
RANDOM_SEED = 42 | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
DatasetLet's download dataset from given [url](https://archive.ics.uci.edu/ml/machine-learning-databases/00228/smsspamcollection.zip), then take a look at samples. Retrieve dataset | def get_dataset():
resp = urlopen("https://archive.ics.uci.edu/ml/machine-learning-databases/00228/smsspamcollection.zip")
zipfile = ZipFile(BytesIO(resp.read()))
lines = list()
for line in zipfile.open('SMSSpamCollection').readlines():
lines.append(line.decode('utf-8'))
data = pd.... | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Explore Samples | dataset.head()
dataset.shape | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Generate Train/Test splits and move forward We see the imbalance in target variable | dataset.label.value_counts() | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
We have duplicated records | dataset.duplicated().sum() | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
remove these duplicates | dataset.drop_duplicates(inplace=True)
dataset.reset_index(drop=True, inplace=True) | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
split into train/test splits with 20/80 ratio | train, test = train_test_split(dataset, test_size=0.2, random_state=RANDOM_SEED) | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Store these sets into dataset directory | DATASET_NAME = "SMSSpamCollection"
if not os.path.exists('data'):
os.mkdir('data')
if not os.path.exists(f'data/{DATASET_NAME}'):
os.mkdir(f'data/{DATASET_NAME}')
train.to_csv(f'data/{DATASET_NAME}/train.csv')
test.to_csv(f'data/{DATASET_NAME}/test.csv') | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Load again and continue | train = pd.read_csv(f'data/{DATASET_NAME}/train.csv', index_col=0)
test = pd.read_csv(f'data/{DATASET_NAME}/test.csv', index_col=0)
train.shape, test.shape
train.head(2) | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Generate EmbeddingsWe use spacy embeddings to vectorize our samples | class Vectorizer:
"""Generates text embedding using deep learning model"""
def __init__(self, *args, **kwargs):
self.model = spacy_sentence_bert.load_model(kwargs.get('model', 'en_paraphrase_distilroberta_base_v1'))
def __call__(self, text):
if not text:
text = ""
... | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
PyTorch ML Pipeline ModelExample of model is taken from [here](https://github.com/rmunro/pytorch_active_learning/blob/master/active_learning_basics.py) | class MLP(nn.Module):
"""Simple 2 Layer Fully Connected NN (MLP)"""
def __init__(self, num_labels, emb_dim):
super(MLP, self).__init__()
# Define model with one hidden layer with 128 neurons
self.linear1 = nn.Linear(emb_dim, 128)
self.linear2 = nn.Linear(128, num_labels)
... | _____no_output_____ | MIT | notebooks/regular_sentiment_analysis_pipeline.ipynb | AnzorGozalishvili/active_learning_playground |
Bias Removal Climate models can have biases relative to different verification datasets. Commonly, biases are removed by postprocessing before verification of forecasting skill. `climpred` provides convenience functions to do so. | import climpred
import xarray as xr
import matplotlib.pyplot as plt
from climpred import HindcastEnsemble
hind = climpred.tutorial.load_dataset('CESM-DP-SST') # CESM-DPLE hindcast ensemble output.
obs = climpred.tutorial.load_dataset('ERSST') # observations
hind["lead"].attrs["units"] = "years" | _____no_output_____ | MIT | docs/source/bias_removal.ipynb | andersy005/climpred |
We begin by removing a mean climatology for the observations, since `CESM-DPLE` generates its anomalies over this same time period. | obs = obs - obs.sel(time=slice('1964', '2014')).mean('time')
hindcast = HindcastEnsemble(hind)
hindcast = hindcast.add_observations(obs)
hindcast.plot() | /Users/rileybrady/Desktop/dev/climpred/climpred/utils.py:141: UserWarning: Assuming annual resolution due to numeric inits. Change init to a datetime if it is another resolution.
"Assuming annual resolution due to numeric inits. "
| MIT | docs/source/bias_removal.ipynb | andersy005/climpred |
The warming of the `observations` is similar to `initialized`. Mean bias removal Typically, bias depends on lead-time and therefore should therefore also be removed depending on lead-time. | bias = hindcast.verify(metric='bias', comparison='e2o', dim=[], alignment='same_verifs')
bias.SST.plot() | _____no_output_____ | MIT | docs/source/bias_removal.ipynb | andersy005/climpred |
Against `observations`, there is small cold bias in 1980 and 1990 initialization years and warm bias before and after. | # lead-time dependant mean bias over all initializations is quite small but negative
mean_bias = bias.mean('init')
mean_bias.SST.plot() | _____no_output_____ | MIT | docs/source/bias_removal.ipynb | andersy005/climpred |
Cross ValidatationTo remove the mean bias quickly, the mean bias over all initializations is subtracted. For formally correct bias removal with cross validation, the given initialization is left out when subtracting the mean bias. `climpred` wraps these functions in `HindcastEnsemble.remove_bias(how='mean', cross_vali... | hindcast.remove_bias(how='mean', cross_validate=True, alignment='same_verifs').plot()
plt.title('hindcast lead timeseries removed for unconditional mean bias')
plt.show() | _____no_output_____ | MIT | docs/source/bias_removal.ipynb | andersy005/climpred |
Skill Distance-based accuracy metrics like (`mse`,`rmse`,`nrmse`,...) are sensitive to mean bias removal. Correlations like (`pearson_r`, `spearman_r`) are insensitive to bias correction. | metric='rmse'
hindcast.verify(metric=metric,
comparison='e2o',
dim='init',
alignment='same_verifs')['SST'].plot(label='no bias correction')
hindcast.remove_bias(cross_validate=False,
alignment='same_verifs') \
.verify(metric=metric,
... | _____no_output_____ | MIT | docs/source/bias_removal.ipynb | andersy005/climpred |
ๆฉๆขฐๅญฆ็ฟใฉใคใใฉใชๆฉๆขฐๅญฆ็ฟใฉใคใใฉใชใKerasใจPyTorchใฎใณใผใใ็ดนไปใใพใใ ไปๅใฏใณใผใใฎ่ฉณใใ่งฃ่ชฌใฏ่กใใพใใใใๅฎ่ฃ
ใฎๅคงใพใใชๆตใใๆๆกใใพใใใใ โ KerasใฎใณใผใไปฅไธใฎใณใผใใฏใKerasใซใใใทใณใใซใชใใฅใผใฉใซใใใใฏใผใฏใฎๅฎ่ฃ
ใงใใ Irisใฎๅ่ฑใใSetosaใจVersicolorใซๅ้กใใพใใ ไปฅไธใฎใณใผใใงใฏใ`Sequential`ใงใขใใซใไฝใใๅฑคใๆดปๆงๅ้ขๆฐใ่ฟฝๅ ใใฆใใพใใ | import numpy as np
from sklearn import datasets
from sklearn import preprocessing
from sklearn.model_selection import train_test_split
iris = datasets.load_iris()
iris_data = iris.data
sl_data = iris_data[:100, 0] # SetosaใจVersicolorใSepal length
sw_data = iris_data[:100, 1] # SetosaใจVersicolorใSepal width
# ๅนณๅๅคใ0ใซ
... | _____no_output_____ | MIT | section_7/ml_libraries.ipynb | kiyoshion/minna-ai |
โ PyTorchใฎใณใผใไปฅไธใฎใณใผใใฏใPyTorchใใใทใณใใซใชใใฅใผใฉใซใใใใฏใผใฏใฎๅฎ่ฃ
ใงใใ Irisใฎๅ่ฑใใSetosaใจVersicolorใซๅ้กใใพใใ ไปฅไธใฎใณใผใใงใฏใKerasใจๅๆงใซ`Sequential`ใงใขใใซใไฝใใๅฑคใๆดปๆงๅ้ขๆฐใไธฆในใฆใใพใใ PyTorchใงใฏใๅ
ฅๅใๆญฃ่งฃใTensorๅฝขๅผใฎใใผใฟใซๅคๆใใๅฟ
่ฆใใใใพใใ | import numpy as np
from sklearn import datasets
from sklearn import preprocessing
from sklearn.model_selection import train_test_split
iris = datasets.load_iris()
iris_data = iris.data
sl_data = iris_data[:100, 0] # SetosaใจVersicolorใSepal length
sw_data = iris_data[:100, 1] # SetosaใจVersicolorใSepal width
# ๅนณๅๅคใ0ใซ
... | _____no_output_____ | MIT | section_7/ml_libraries.ipynb | kiyoshion/minna-ai |
Transfer Learning Template | %load_ext autoreload
%autoreload 2
%matplotlib inline
import os, json, sys, time, random
import numpy as np
import torch
from torch.optim import Adam
from easydict import EasyDict
import matplotlib.pyplot as plt
from steves_models.steves_ptn import Steves_Prototypical_Network
from steves_utils.lazy_iterable_wr... | _____no_output_____ | MIT | experiments/tl_2/cores_wisig-oracle.run2.framed/trials/1/trial.ipynb | stevester94/csc500-notebooks |
Allowed ParametersThese are allowed parameters, not defaultsEach of these values need to be present in the injected parameters (the notebook will raise an exception if they are not present)Papermill uses the cell tag "parameters" to inject the real parameters below this cell.Enable tags to see what I mean | required_parameters = {
"experiment_name",
"lr",
"device",
"seed",
"dataset_seed",
"n_shot",
"n_query",
"n_way",
"train_k_factor",
"val_k_factor",
"test_k_factor",
"n_epoch",
"patience",
"criteria_for_best",
"x_net",
"datasets",
"torch_default_dtype",
... | _____no_output_____ | MIT | experiments/tl_2/cores_wisig-oracle.run2.framed/trials/1/trial.ipynb | stevester94/csc500-notebooks |
Supplementary | def show_results_dist_1_3(frames, title, y_axe_name,
dims=(18, 12), save=False, file_name='trash', legend_size=13):
size = len(frames)
a4_dims = dims
fig, axs = plt.subplots(1, 3, figsize=a4_dims)
for i in range(3):
sns.lineplot(x="Error = 1 - Recall@1", y=y_axe_name,hu... | _____no_output_____ | MIT | results/draw_results.ipynb | Shekhale/gbnns_theory |
y = 0.14766146x + 12.10283073 | lr.score(x,y) #์ ํ๋ | _____no_output_____ | Apache-2.0 | autompg_regression-Copy1.ipynb | power3247/test_machinlearning |
์ฌ์ดํท ๋ฐ ๋ชจ๋ธ๋ง์ ํตํด์ ๋ ๋ฐ์ดํฐ๋ฅผ ์ฃผ์ด์ฃผ๊ณ ์ ํ๋ชจ๋ธ์ ๋ง๋ ๊ฒ | y_predicted = lr.predict([[99]])
y_predicted | _____no_output_____ | Apache-2.0 | autompg_regression-Copy1.ipynb | power3247/test_machinlearning |
1. ์ ๋ณด๋จ๊ณ: ์์ง ๊ฐ๊ณต- ๋ฌธ์ ๋ฐ์ดํฐ ํ์ธ ๋ฐ ์ฒ๋ฆฌ2. ๊ต์ก ๋จ๊ณ: ๋จธ์ ๋์- ์ปฌ๋ผ์ ํ- ๋ชจ๋ธ ์ ํ- ๊ต์ก- ์ ํ๋ ํ์ธ3. ์๋น์ค ๋จ๊ณ:๊ณ ๊ฐ์๋ X์ ๊ฐ์ ๋๋ฆด์ ์๋ค | from sklearn.model_selection import train_test_split
X_train, X_test, Y_train, Y_test = train_test_split(x, y)
X_train, X_test, Y_train, Y_test
lr.fit(X_train, Y_train)
lr.coef_, lr.intercept_ #๊ธฐ์ธ๊ธฐ, ์ ํธ
lr.score(X_train, Y_train)
lr.score(X_test,Y_test)
lr.predict([[3504.0, 8]])
lr.predict([[2790.0, 4]])
import pickle
p... | _____no_output_____ | Apache-2.0 | autompg_regression-Copy1.ipynb | power3247/test_machinlearning |
Fig.3 - Comparisons of MDR's AUC across 12 different sets Change plot's default size and font | import matplotlib.pyplot as plt
plt.rcParams['figure.figsize'] = [20, 12]
rc = {"font.family" : "sans-serif",
"font.style" : "normal",
"mathtext.fontset" : "dejavusans"}
plt.rcParams.update(rc)
plt.rcParams["font.sans-serif"] = ["Myriad Pro"] + plt.rcParams["font.sans-serif"]
from constant import HEADER_NA... | _____no_output_____ | MIT | mdr_analysis/3Fig-version0-AUC_IQR_across_therapy_and_sets-separate.ipynb | bonilab/malariaibm-generation-of-MDR-mutants |
PfPR = 0.1% | fig, axes = plt.subplots(5, 3, sharex='col', sharey='row',
gridspec_kw={'hspace': 0, 'wspace': 0})
fig.patch.set_facecolor('white')
fig.suptitle('PfPR = 0.1%', y=0.93, fontweight='bold', fontsize=TITLE_FONTSIZE)
# plot IQR of AUC for most-dangerous-triple and -double
from plotter import fig3_da... | _____no_output_____ | MIT | mdr_analysis/3Fig-version0-AUC_IQR_across_therapy_and_sets-separate.ipynb | bonilab/malariaibm-generation-of-MDR-mutants |
PfPR = 1% | fig, axes = plt.subplots(5, 3, sharex='col', sharey='row',
gridspec_kw={'hspace': 0, 'wspace': 0})
fig.patch.set_facecolor('white')
fig.suptitle('PfPR = 1%', y=0.93, fontweight='bold', fontsize=TITLE_FONTSIZE)
# plot IQR of AUC for most-dangerous-triple and -double
from plotter import fig3_dang... | _____no_output_____ | MIT | mdr_analysis/3Fig-version0-AUC_IQR_across_therapy_and_sets-separate.ipynb | bonilab/malariaibm-generation-of-MDR-mutants |
PfPR = 5% | fig, axes = plt.subplots(5, 3, sharex='col', sharey='row',
gridspec_kw={'hspace': 0, 'wspace': 0})
fig.patch.set_facecolor('white')
fig.suptitle('PfPR = 5%', y=0.93, fontweight='bold', fontsize=TITLE_FONTSIZE)
# plot IQR of AUC for most-dangerous-triple and -double
from plotter import fig3_dang... | _____no_output_____ | MIT | mdr_analysis/3Fig-version0-AUC_IQR_across_therapy_and_sets-separate.ipynb | bonilab/malariaibm-generation-of-MDR-mutants |
PfPR = 20% | fig, axes = plt.subplots(5, 3, sharex='col', sharey='row',
gridspec_kw={'hspace': 0, 'wspace': 0})
fig.patch.set_facecolor('white')
fig.suptitle('PfPR = 20%', y=0.93, fontweight='bold', fontsize=TITLE_FONTSIZE)
# plot IQR of AUC for most-dangerous-triple and -double
from plotter import fig3_dan... | _____no_output_____ | MIT | mdr_analysis/3Fig-version0-AUC_IQR_across_therapy_and_sets-separate.ipynb | bonilab/malariaibm-generation-of-MDR-mutants |
Image ClassificationIn this project, you'll classify images from the [CIFAR-10 dataset](https://www.cs.toronto.edu/~kriz/cifar.html). The dataset consists of airplanes, dogs, cats, and other objects. You'll preprocess the images, then train a convolutional neural network on all the samples. The images need to be norm... | """
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
from urllib.request import urlretrieve
from os.path import isfile, isdir
from tqdm import tqdm
import problem_unittests as tests
import tarfile
cifar10_dataset_folder_path = 'cifar-10-batches-py'
# Use Floyd's cifar-10 dataset if present
floyd_cifar10... | CIFAR-10 Dataset: 171MB [01:58, 1.44MB/s]
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Explore the DataThe dataset is broken into batches to prevent your machine from running out of memory. The CIFAR-10 dataset consists of 5 batches, named `data_batch_1`, `data_batch_2`, etc.. Each batch contains the labels and images that are one of the following:* airplane* automobile* bird* cat* deer* dog* frog* hor... | %matplotlib inline
%config InlineBackend.figure_format = 'retina'
import helper
import numpy as np
# Explore the dataset
batch_id = 1
sample_id = 5
helper.display_stats(cifar10_dataset_folder_path, batch_id, sample_id) |
Stats of batch 1:
Samples: 10000
Label Counts: {0: 1005, 1: 974, 2: 1032, 3: 1016, 4: 999, 5: 937, 6: 1030, 7: 1001, 8: 1025, 9: 981}
First 20 Labels: [6, 9, 9, 4, 1, 1, 2, 7, 8, 3, 4, 7, 7, 2, 9, 9, 9, 3, 2, 6]
Example of Image 5:
Image - Min Value: 0 Max Value: 252
Image - Shape: (32, 32, 3)
Label - Label Id: 1 Nam... | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Implement Preprocess Functions NormalizeIn the cell below, implement the `normalize` function to take in image data, `x`, and return it as a normalized Numpy array. The values should be in the range of 0 to 1, inclusive. The return object should be the same shape as `x`. | def normalize(x):
"""
Normalize a list of sample image data in the range of 0 to 1
: x: List of image data. The image shape is (32, 32, 3)
: return: Numpy array of normalize data
"""
min = np.amin(x)
max = np.amax(x)
return (x - min) / (max - min)
"""
DON'T MODIFY ANYTHING IN THIS CEL... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
One-hot encodeJust like the previous code cell, you'll be implementing a function for preprocessing. This time, you'll implement the `one_hot_encode` function. The input, `x`, are a list of labels. Implement the function to return the list of labels as One-Hot encoded Numpy array. The possible values for labels are... | x = np.array([6, 1, 5])
np.array([[1 if i == y else 0 for i in range(10)] for y in x])
def one_hot_encode(x):
"""
One hot encode a list of sample labels. Return a one-hot encoded vector for each label.
: x: List of sample Labels
: return: Numpy array of one-hot encoded labels
"""
return np.array... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Randomize DataAs you saw from exploring the data above, the order of the samples are randomized. It doesn't hurt to randomize it again, but you don't need to for this dataset. Preprocess all the data and save itRunning the code cell below will preprocess all the CIFAR-10 data and save it to file. The code below also... | """
DON'T MODIFY ANYTHING IN THIS CELL
"""
# Preprocess Training, Validation, and Testing Data
helper.preprocess_and_save_data(cifar10_dataset_folder_path, normalize, one_hot_encode) | _____no_output_____ | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Check PointThis is your first checkpoint. If you ever decide to come back to this notebook or have to restart the notebook, you can start from here. The preprocessed data has been saved to disk. | """
DON'T MODIFY ANYTHING IN THIS CELL
"""
import pickle
import problem_unittests as tests
import helper
# Load the Preprocessed Validation data
valid_features, valid_labels = pickle.load(open('preprocess_validation.p', mode='rb')) | _____no_output_____ | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Build the networkFor the neural network, you'll build each layer into a function. Most of the code you've seen has been outside of functions. To test your code more thoroughly, we require that you put each layer in a function. This allows us to give you better feedback and test for simple mistakes using our unittest... | import tensorflow as tf
def neural_net_image_input(image_shape):
"""
Return a Tensor for a batch of image input
: image_shape: Shape of the images
: return: Tensor for image input.
"""
return tf.placeholder(tf.float32, shape = [None] + list(image_shape), name = "x")
def neural_net_label_input... | Image Input Tests Passed.
Label Input Tests Passed.
Keep Prob Tests Passed.
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Convolution and Max Pooling LayerConvolution layers have a lot of success with images. For this code cell, you should implement the function `conv2d_maxpool` to apply convolution then max pooling:* Create the weight and bias using `conv_ksize`, `conv_num_outputs` and the shape of `x_tensor`.* Apply a convolution to `x... | def conv2d_maxpool(x_tensor, conv_num_outputs, conv_ksize, conv_strides, pool_ksize, pool_strides):
"""
Apply convolution then max pooling to x_tensor
:param x_tensor: TensorFlow Tensor
:param conv_num_outputs: Number of outputs for the convolutional layer
:param conv_ksize: kernal size 2-D Tuple fo... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Flatten LayerImplement the `flatten` function to change the dimension of `x_tensor` from a 4-D tensor to a 2-D tensor. The output should be the shape (*Batch Size*, *Flattened Image Size*). Shortcut option: you can use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [Tens... | import numpy as np
def flatten(x_tensor):
"""
Flatten x_tensor to (Batch Size, Flattened Image Size)
: x_tensor: A tensor of size (Batch Size, ...), where ... are the image dimensions.
: return: A tensor of size (Batch Size, Flattened Image Size).
"""
x_size = x_tensor.get_shape().as_list()
... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Fully-Connected LayerImplement the `fully_conn` function to apply a fully connected layer to `x_tensor` with the shape (*Batch Size*, *num_outputs*). Shortcut option: you can use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tens... | def fully_conn(x_tensor, num_outputs):
"""
Apply a fully connected layer to x_tensor using weight and bias
: x_tensor: A 2-D tensor where the first dimension is batch size.
: num_outputs: The number of output that the new tensor should be.
: return: A 2-D tensor where the second dimension is num_out... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Output LayerImplement the `output` function to apply a fully connected layer to `x_tensor` with the shape (*Batch Size*, *num_outputs*). Shortcut option: you can use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tensorflow.org/ap... | def output(x_tensor, num_outputs):
"""
Apply a output layer to x_tensor using weight and bias
: x_tensor: A 2-D tensor where the first dimension is batch size.
: num_outputs: The number of output that the new tensor should be.
: return: A 2-D tensor where the second dimension is num_outputs.
"""... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Create Convolutional ModelImplement the function `conv_net` to create a convolutional neural network model. The function takes in a batch of images, `x`, and outputs logits. Use the layers you created above to create this model:* Apply 1, 2, or 3 Convolution and Max Pool layers* Apply a Flatten Layer* Apply 1, 2, or ... | def conv_net(x, keep_prob):
"""
Create a convolutional neural network model
: x: Placeholder tensor that holds image data.
: keep_prob: Placeholder tensor that hold dropout keep probability.
: return: Tensor that represents logits
"""
# TODO: Apply 1, 2, or 3 Convolution and Max Pool layers
... | Neural Network Built!
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Train the Neural Network Single OptimizationImplement the function `train_neural_network` to do a single optimization. The optimization should use `optimizer` to optimize in `session` with a `feed_dict` of the following:* `x` for image input* `y` for labels* `keep_prob` for keep probability for dropoutThis function w... | def train_neural_network(session, optimizer, keep_probability, feature_batch, label_batch):
"""
Optimize the session on a batch of images and labels
: session: Current TensorFlow session
: optimizer: TensorFlow optimizer function
: keep_probability: keep probability
: feature_batch: Batch of Num... | Tests Passed
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Show StatsImplement the function `print_stats` to print loss and validation accuracy. Use the global variables `valid_features` and `valid_labels` to calculate validation accuracy. Use a keep probability of `1.0` to calculate the loss and validation accuracy. | def print_stats(session, feature_batch, label_batch, cost, accuracy):
"""
Print information about loss and validation accuracy
: session: Current TensorFlow session
: feature_batch: Batch of Numpy image data
: label_batch: Batch of Numpy label data
: cost: TensorFlow cost function
: accuracy... | _____no_output_____ | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
HyperparametersTune the following parameters:* Set `epochs` to the number of iterations until the network stops learning or start overfitting* Set `batch_size` to the highest number that your machine has memory for. Most people set them to common sizes of memory: * 64 * 128 * 256 * ...* Set `keep_probability` to the ... | # TODO: Tune Parameters
epochs = 10
batch_size = 256
keep_probability = .5 | _____no_output_____ | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Train on a Single CIFAR-10 BatchInstead of training the neural network on all the CIFAR-10 batches of data, let's use a single batch. This should save time while you iterate on the model to get a better accuracy. Once the final validation accuracy is 50% or greater, run the model on all the data in the next section. | """
DON'T MODIFY ANYTHING IN THIS CELL
"""
print('Checking the Training on a Single Batch...')
with tf.Session() as sess:
# Initializing the variables
sess.run(tf.global_variables_initializer())
# Training cycle
for epoch in range(epochs):
batch_i = 1
for batch_features, batch_label... | Checking the Training on a Single Batch...
Epoch 1, CIFAR-10 Batch 1: Training Loss= 2.303360, Training Accuracy= 0.07500
Validation Loss= 2.302445, Validation Accuracy= 0.10020
Epoch 2, CIFAR-10 Batch 1: Training Loss= 2.300095, Training Accuracy= 0.07500
Validation Loss= 2.300672, Validation Accuracy= 0.10680
Epo... | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Fully Train the ModelNow that you got a good accuracy with a single CIFAR-10 batch, try it with all five batches. | """
DON'T MODIFY ANYTHING IN THIS CELL
"""
save_model_path = './image_classification'
print('Training...')
with tf.Session() as sess:
# Initializing the variables
sess.run(tf.global_variables_initializer())
# Training cycle
for epoch in range(epochs):
# Loop over all batches
n_batc... | Training...
Epoch 1, CIFAR-10 Batch 1: Training Loss= 2.267260, Training Accuracy= 0.10000
Validation Loss= 2.247928, Validation Accuracy= 0.15120
Epoch 1, CIFAR-10 Batch 2: Training Loss= 1.995992, Training Accuracy= 0.17500
Validation Loss= 2.067119, Validation Accuracy= 0.22640
Epoch 1, CIFAR-10 Batch 3: Train... | MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
CheckpointThe model has been saved to disk. Test ModelTest your model against the test dataset. This will be your final accuracy. You should have an accuracy greater than 50%. If you don't, keep tweaking the model architecture and parameters. | """
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
%config InlineBackend.figure_format = 'retina'
import tensorflow as tf
import pickle
import helper
import random
# Set batch size if not already set
try:
if batch_size:
pass
except NameError:
batch_size = 64
save_model_path = './image_clas... | Testing Accuracy: 0.5068359375
| MIT | image-classification/dlnd_image_classification.ipynb | mdiaz236/DeepLearningFoundations |
Least-squares technique References - Statistics in geography: https://archive.org/details/statisticsingeog0000ebdo/ Imports | from functools import partial
import numpy as np
from scipy.stats import multivariate_normal, t
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from ipywidgets import interact, IntSlider
inv = np.linalg.inv
df = pd.read_csv('regression_data.csv')
df.head(3) | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Population 0.5 and 0.2 are NOT the population parameters. Although we used them to generate the population, the population parameters can be different from them. | def get_y(x):
ys = x * 0.5 + 0.2
noises = 1 * np.random.normal(size=len(ys))
return ys + noises
np.random.seed(52)
xs = np.linspace(0, 10, 10000)
ys = get_y(xs)
np.random.seed(32)
np.random.shuffle(xs)
np.random.seed(32)
np.random.shuffle(ys)
plt.scatter(xs, ys, s=5)
plt.show() | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Design matrices | PHI = xs.reshape(-1, 1)
PHI = np.hstack([
PHI,
np.ones(PHI.shape)
])
T = ys.reshape(-1, 1) | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Normal equation with regularization | def regularized_least_squares(PHI, T, regularizer=0):
assert PHI.shape[0] == T.shape[0]
pseudo_inv = inv(PHI.T @ PHI + np.eye(PHI.shape[1]) * regularizer)
assert pseudo_inv.shape[0] == pseudo_inv.shape[1]
W = pseudo_inv @ PHI.T @ T
return {'slope' : float(W[0]), 'intercept' ... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Sampling distributions Population parameters | pop_params = regularized_least_squares(PHI, T)
pop_slope, pop_intercept = pop_params['slope'], pop_params['intercept'] | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Sample statistics Verify that the sampling distribution for both regression coefficients are normal. | n = 10 # sample size
num_samps = 1000
def sample(PHI, T, n):
idxs = np.random.randint(PHI.shape[0], size=n)
return PHI[idxs], T[idxs]
samp_slopes, samp_intercepts = [], []
for i in range(num_samps):
PHI_samp, T_samp = sample(PHI, T, n)
learned_param = regularized_least_squares(PHI_samp, T_samp)
sam... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Note that the two normal distributions above are correlated. This means that we need to be careful when plotting the 95% CI for the regression line, because we can't just plot the regression line with the highest slope and the highest intercept and the regression line with the lowest slope and the lowest intercept. | sns.jointplot(samp_slopes, samp_intercepts, s=5)
plt.show() | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Confidence interval **Caution.** The following computation of confidence intervals does not apply to regularized least squares. Sample one sample | n = 500
PHI_samp, T_samp = sample(PHI, T, n) | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Compute sample statistics | learned_param = regularized_least_squares(PHI_samp, T_samp)
samp_slope, samp_intercept = learned_param['slope'], learned_param['intercept']
samp_slope, samp_intercept | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Compute standard errors of sample statisticsStandard error is the estimate of the standard deviation of the sampling distribution. $$\hat\sigma = \sqrt{\frac{\text{Sum of all squared residuals}}{\text{Degrees of freedom}}}$$ Standard error for slope:$$\text{SE}(\hat\beta_1)=\hat\sigma \sqrt{\frac{1}{(n-1)s_X^2}}$$ Sta... | preds = samp_slope * PHI_samp[:,0] + samp_intercept
sum_of_squared_residuals = np.sum((T_samp.reshape(-1) - preds) ** 2)
samp_sigma_y_give_x = np.sqrt(sum_of_squared_residuals / (n - 2))
samp_sigma_y_give_x
samp_mean = np.mean(PHI_samp[:,0])
samp_var = np.var(PHI_samp[:,0])
SE_slope = samp_sigma_y_give_x * np.sqrt(1 / ... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Compute confidence intervals for sample statistics | slope_lower, slope_upper = samp_slope - 1.96 * SE_slope, samp_slope + 1.96 * SE_slope
slope_lower, slope_upper
intercept_lower, intercept_upper = samp_intercept - 1.96 * SE_intercept, samp_intercept + 1.96 * SE_intercept
intercept_lower, intercept_upper | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Compute confidence interval for regression line Boostrapped solution Use a 2-d Guassian to model the joint distribution between boostrapped sample slopes and boostrapped sample intercepts. **Fixed.** `samp_slopes` and `samp_intercepts` used in the cell below are not boostrapped; they are directly sampled from the pop... | num_resamples = 10000
resample_slopes, resample_intercepts = [], []
for i in range(num_resamples):
PHI_resample, T_resample = sample(PHI_samp, T_samp, n=len(PHI_samp))
learned_params = regularized_least_squares(PHI_resample, T_resample)
resample_slopes.append(learned_params['slope']); resample_intercepts.ap... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
**Fixed.** The following steps might improve the results, but I don't think they are part of the standard practice. | # means = [np.mean(resample_slopes), np.mean(resample_intercepts)]
# cov = np.cov(resample_slopes, resample_intercepts)
# model = multivariate_normal(mean=means, cov=cov) | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Sample 5000 (slope, intercept) pairs from the Gaussian. | # num_pairs_sampled = 10000
# pairs = model.rvs(num_pairs_sampled) | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Scatter samples, plot regression lines and CI. | plt.figure(figsize=(20, 10))
plt.scatter(PHI_samp[:,0], T_samp.reshape(-1), s=20) # sample
granularity = 1000
xs = np.linspace(0, 10, granularity)
plt.plot(xs, samp_slope * xs + samp_intercept, label='Sample') # sample regression line
plt.plot(xs, pop_slope * xs + pop_intercept, '--', color='black', label='Popu... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Analytic solution **Reference.** Page 97, Statistics of Geograph: A Practical Approach, David Ebdon, 1987. For a particular value $x_0$ of the independent variable $x$, its confidence interval is given by: $$\sqrt{\frac{\sum e^{2}}{n-2}\left[\frac{1}{n}+\frac{\left(x_{0}-\bar{x}\right)^{2}}{\sum x^{2}-n \bar{x}^{2}}\r... | sum_of_squared_xs = np.sum(PHI_samp[:,0] ** 2)
SEs = np.sqrt(
(sum_of_squared_residuals / (n - 2)) *
(1 / n + (xs - samp_mean) ** 2 / (sum_of_squared_xs - n * samp_mean ** 2))
)
t_97dot5 = t.ppf(0.975, df=n-2)
t_99dot5 = t.ppf(0.995, df=n-2)
yhats = samp_slope * xs + samp_intercept
uppers_95 = yhats + t_97dot5... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Regularized least squares | def plot_regression_line(PHI, T, regularizer):
plt.scatter(PHI[:,0], T, s=5)
params = regularized_least_squares(PHI, T, regularizer)
x_min, x_max = PHI[:,0].min(), PHI[:,0].max()
xs = np.linspace(x_min, x_max, 2)
ys = params['slope'] * xs + params['intercept']
plt.plot(xs, ys... | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
Yes! The effect of regularization does change with the size of the dataset. | _ = interact(
plot_regression_line_wrapper,
regularizer=IntSlider(min=0, max=10000, value=5000, continuous_update=False),
num_points=IntSlider(min=2, max=1000, value=1000, continuous_update=False)
) | _____no_output_____ | MIT | c3_single_layer_networks/least_square_method/least_square_method_demo.ipynb | zhihanyang2022/bishop1995_notes |
1. ์ ๊ท ํํ์* ์ถ์ฒ : ์์ "์ก์๋ผ! ํ
์คํธ ๋ง์ด๋ with ํ์ด์ฌ"* ๋ฌธ์์ด์ด ์ฃผ์ด์ง ๊ท์น์ ์ผ์นํ๋ ์ง, ์ผ์นํ์ง ์๋์ง ํ๋จํ ์ ์๋ค. ์ ๊ท ํํ์์ ์ด์ฉํ์ฌ ํน์ ํจํด์ ์ง๋๋ ๋ฌธ์์ด์ ์ฐพ์ ์ ์์ด ํ
์คํธ ๋ฐ์ดํฐ ์ฌ์ ์ฒ๋ฆฌ ๋ฐ ํฌ๋กค๋ง์์ ์ฃผ๋ก ์ฐ์ | string = '๊ธฐ์์ฒญ์ ์ํผ์ปดํจํฐ๋ ์์ธ์ง์ญ์ ์ง์คํธ์ฐ๋ฅผ ์ ๋๋ก ์์ธกํ์ง ๋ชปํ๋ค๊ณ ์ค๋ช
ํ์ต๋๋ค. ์ ์ค๋ฅ๊ฐ ๋ฐ์ํ\
์ต๋๋ค. ์ ์ค๋ฅ๊ฐ ๋ฐ์ํ๋์ง ์์ธํ ๋ถ์ํด ์์ธก ํ๋ก๊ทธ๋จ์ ๋ณด์ํด์ผ ํ ๋๋ชฉ์
๋๋ค. ๊ด์ธก ๋ถ์ผ๋ ๊ฐ์ ๋ ์ฌ์ง๊ฐ\
์์ต๋๋ค. ์ง๊ธ ๋ณด์๋ ์ผ์ชฝ ์ฌ์ง์ด ํ์ฌ ์ฒ๋ฆฌ์ ์์ฑ์ด ์ดฌ์ํ ๊ฒ์ด๊ณ ์ค๋ฅธ์ชฝ์ด ์ฌํด ๋ง ์์ ์ฌ๋ฆด ์ฒ๋ฆฌ์ 2Aํธ๊ฐ \
์ดฌ์ํ ์์์
๋๋ค. ์ค๋ฅธ์ชฝ์ด ์ผ์ชฝ๋ณด๋ค ํํ์ ๋์ด ์ข ๋ ๋๋ ทํ๊ณ ์ฃผ๋ณ ๊ตฌ๋ฆ๋ ๋ ์ ๋ณด์ด์ฃ . ์ ์์ฑ์ ํตํด ํํ\
๊ตฌ๋ฆ ๋ฑ์ ์์ง์์ ์์ธํ ๋ถ์ํ๋ฉด ์ข ๋ ์ ํํ ์์ธก์ ํ ์ ์์ง ์์๊น ๊ธฐ๋ํด ๋ด
๋๋ค. ์ ๊ตฌํฌ ... | _____no_output_____ | MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
* \([A-Za-z0-9\._+]+ : ์ด๋ฉ์ผ ์ฃผ์๊ฐ ๊ดํธ๋ก ์์ํ์ฌ \(ํน์๋ฌธ์๋ฅผ ์๋ ์๋ฏธ๋๋ก ์ฐ๊ฒ ํจ)์ (๋ก ์์ ๋๊ดํธ[ ] ์์ ์ด๋ฉ์ผ ์ฃผ์์ ํจํด์ ์
๋ ฅ(์
๋ ฅํ ๊ฒ ์ค ์๋ฌด๊ฑฐ๋) A-Z = ์ํ๋ฒณ ๋๋ฌธ์, a-z = ์ํ๋ฒณ ์๋ฌธ์, 0-9 = ์ซ์, ._+ = .๋ _๋ + ๋ง์ง๋ง +๋ ๋ฐ๋ก ์์ ์๋ ๊ฒ์ด ์ต์ ํ๋ฒ ์ด์ ๋์์ผ ํ๋ค๋ ์๋ฏธ * @ : ์ด๋ฉ์ผ ์ฃผ์ ๋ค์์ @ * [A-Za-z]+ : ๋๋ฉ์ธ ์ฃผ์์... | # a๋ฌธ์๊ฐ 1๋ฒ ์ด์ ๋์ค๊ณ b ๋ฌธ์๊ฐ 0๋ฒ ์ด์ ๋์ค๋ ๋ฌธ์์ด ์ฐพ๊ธฐ
r = re.compile("a+b*")
r.findall("aaaa, cc, bbbb, aabbb")
# ๋๊ดํธ๋ฅผ ์ด์ฉํด ๋๋ฌธ์๋ก ๊ตฌ์ฑ๋ ๋ฌธ์์ด ์ฐพ๊ธฐ
r = re.compile("[A-Z]+")
r.findall("HOME, home")
# ^์ .์ ์ด์ฉํ์ฌ ๋งจ ์์ a๊ฐ ์ค๊ณ ๊ทธ ๋ค์์ ์ด๋ ํ ํํ๋ 2๊ฐ์ ๋ฌธ์๊ฐ ์ค๋ ๋ฌธ์์ด ์ฐพ๊ธฐ
r = re.compile("^a..")
r.findall("abc, cba")
# ์ค๊ดํธ ํํ์ {m,n}์ ์ด์ํ์ฌ ํด๋น ๋ฌธ์์ด์ด m๋ฒ ์ด์ n๋ฒ ์ดํ ๋ํ๋... | _____no_output_____ | MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
2. ์ ์ฒ๋ฆฌ ๋์๋ฌธ์ ํต์ผ | s = 'Hello World'
print(s.lower())
print(s.upper()) | hello world
HELLO WORLD
| MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
์ซ์, ๋ฌธ์ฅ๋ถํธ, ํน์๋ฌธ์ ์ ๊ฑฐ | # ์ซ์ ์ ๊ฑฐ
p = re.compile("[0-9]+")
p.sub("", '์์ธ ๋ถ๋์ฐ ๊ฐ๊ฒฉ์ด ์ฌํด ๋ค์ด ํ๊ท 30% ์์นํ์ต๋๋ค.')
# ๋ฌธ์ฅ๋ถํธ, ํน์๋ฌธ์ ์ ๊ฑฐ
p = re.compile("\W+")
p.sub(" ", "โ
์์ธ ๋ถ๋์ฐ ๊ฐ๊ฒฉ์ด ์ฌํด ๋ค์ด ํ๊ท 30% ์์นํ์ต๋๋ค.")
s = p.sub(" ", "์ฃผ์ _1: ๊ฑด๊ฐํ ๋ฌผ๊ณผ ๊ฑด๊ฐํ ์ ์ !")
s
p = re.compile("_")
p.sub(" ", s) | _____no_output_____ | MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
๋ถ์ฉ์ด ์ ๊ฑฐ | words_Korean = ['์ถ์','์ฐํด','๋ฏผ์กฑ','๋์ด๋','์์','๋์ด','๊ตํต๋','๊ตํต์ฌ๊ณ ','ํนํ','์๋์ฐจ','๊ณ ์ฅ',
'์๋น์','์ฐจ์ง','๋ํ','๊ฒ','๊ธฐ์']
stopwords = ['๊ฐ๋ค','๋์ด','๋ํ','๊ฒ','๊ธฐ์']
[i for i in words_Korean if i not in stopwords]
from nltk.corpus import stopwords
# ๊ทธ๋ฅํ๋ฉด LookupError ๋ฐ์ํ๋ฏ๋ก ๋ค์ด๋ก๋๊ฐ ํ์ํจ
# import nltk
# nltk.download() or nltk.downlo... | _____no_output_____ | MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
๊ฐ์ ์ด๊ทผ ๋์ผํ(stemming) | from nltk.tokenize import word_tokenize
from nltk.stem import PorterStemmer
ps_stemmer = PorterStemmer()
new_text = 'It is important to be immersed while you are pythoning with python.\
All pythoners have pothoned poorly at least once.'
words = word_tokenize(new_text)
for w in words:
print(ps_stemmer.stem(w), e... | It is important to be immersed while you are ing with .All ers have pothoned poorly at least once . | MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
N-gram* n๋ฒ ์ฐ์ด์ด ๋ฑ์ฅํ๋ ๋จ์ด๋ค์ ์ฐ์* ๋ ๋ฒ : ๋ฐ์ด๊ทธ๋จ, ์ธ ๋ฒ : ํธ๋ผ์ด๊ทธ๋จ(ํธ๋ผ์ด๊ทธ๋จ ์ด์์ ๋ณดํธ์ ์ผ๋ก ํ์ฉํ์ง ์์) | from nltk import ngrams
sentence = 'Chief Justice Roberts, Preskdent Carter, President Clinton, President Bush, President Obama, \
fellow Americans and people of the world, thank you. We, the citizens of America are now joined in a great \
national effort to rebuild our country and restore its promise for all of our pe... | ('Chief', 'Justice') ('Justice', 'Roberts,') ('Roberts,', 'Preskdent') ('Preskdent', 'Carter,') ('Carter,', 'President') ('President', 'Clinton,') ('Clinton,', 'President') ('President', 'Bush,') ('Bush,', 'President') ('President', 'Obama,') ('Obama,', 'fellow') ('fellow', 'Americans') ('Americans', 'and') ('and', 'pe... | MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
ํ์ฌ ๋ถ์ | from konlpy.tag import Hannanum
hannanum = Hannanum()
print(hannanum.morphs("์น์ฒ๋ค์ด ๋ชจ์ธ ์ด๋ฒ ์ถ์ ์ฐจ๋ก์์์๋ ๋จ์ฐ '์ง๊ฐ'์ด ํ์ ์ ์ฌ๋๋ค."))
print(hannanum.nouns("์น์ฒ๋ค์ด ๋ชจ์ธ ์ด๋ฒ ์ถ์ ์ฐจ๋ก์์์๋ ๋จ์ฐ '์ง๊ฐ'์ด ํ์ ์ ์ฌ๋๋ค."))
print(hannanum.pos("์น์ฒ๋ค์ด ๋ชจ์ธ ์ด๋ฒ ์ถ์ ์ฐจ๋ก์์์๋ ๋จ์ฐ '์ง๊ฐ'์ด ํ์ ์ ์ฌ๋๋ค.", ntags=9))
from konlpy.tag import Kkma
kkma = Kkma()
print(kkma.morphs("์น์ฒ... | [('The', 'DT'), ('little', 'JJ'), ('yellow', 'JJ'), ('dog', 'NN'), ('barked', 'VBD'), ('at', 'IN'), ('the', 'DT'), ('Persian', 'NNP'), ('cat.', 'NN')]
| MIT | [STUDY]ํ
์คํธ๋ง์ด๋_Base.ipynb | kamzzang/ADPStudy |
Location-based algorithm | location_map = list(set(TRAINING_DATA_GRAD["LOCATION"]))
location_map.sort()
# print(location_map)
list_location = []
list_avg_rm = []
list_avg_yield = []
for val in location_map:
avg_rm = np.average(TRAINING_DATA_GRAD[EXPERIMENT_DATA["LOCATION"] == str(val)]["RM"])
avg_yield = np.average(TRAINING_DATA_GRAD[EXP... | _____no_output_____ | MIT | Average_RM_YIELD_location.ipynb | hmc-cs-nsuaysom/BigDataProject |
AnalysisFrom the preliminary analysis, we find that the number of different locateion in the dataset is 140. The location in the dataset is encoded as a 4-digit number. We first expected that we can group the quality of the species based on the location parameters. We then plot the average of __RM__ and __YIELD__ for ... | # Calculate the number of possible locations
location_set = set(TRAINING_DATA_GRAD["LOCATION"])
print("The number of possible location is {}.".format(len(location_set)))
location_histogram_list = []
for location in sorted(location_set):
amount = len(TRAINING_DATA_GRAD[TRAINING_DATA_GRAD["LOCATION"] == str(location... | 1.05328277023 1.0262956544
0.762338198505 0.873119807647
0.253807203837 0.503792818366
0.200542622706 0.447819855195
0.29628093697 0.544316945327
0.374676270693 0.612108054753
3.15544362088e-30 1.7763568394e-15
| MIT | Average_RM_YIELD_location.ipynb | hmc-cs-nsuaysom/BigDataProject |
Test with validation set | # Test with validation set
TESTING_DATA_GRAD = TESTING_DATA_GRAD.reset_index(drop=True)
Xtest = np.column_stack((TESTING_DATA_GRAD["LOCATION"],
TESTING_DATA_GRAD["RM"],
TESTING_DATA_GRAD["YIELD"]))
ytest = TESTING_DATA_GRAD["BAGSOLD"].astype(np.float)
log_ytest = np.... | _____no_output_____ | MIT | Average_RM_YIELD_location.ipynb | hmc-cs-nsuaysom/BigDataProject |
Testing Ridge Reg vs. Linear RegBelow is not used. It's for testing the difference between Ridge Regression and Linear Regression.The result is that the MSE is almsot the same | bagSold = np.log(np.asarray(TRAINING_DATA_GRAD["BAGSOLD"]).reshape(-1, 1).astype(np.float))
rm = np.asarray(TRAINING_DATA_GRAD["RM"]).reshape(-1,1).astype(np.float)
yield_val = np.asarray(TRAINING_DATA_GRAD["YIELD"]).reshape(-1,1).astype(np.float)
x = np.column_stack((rm, yield_val))
# Liear Regression
regr = lin... | Coefficients = [[ 0.0062339 0.00430852]]
Intercept = [ 13.11871673]
MSE = 0.3483090550422536
| MIT | Average_RM_YIELD_location.ipynb | hmc-cs-nsuaysom/BigDataProject |
Analyze the range of values from the audio | from scipy.io import wavfile
wav_file = "../data/raw/dadosBruno/t02/" + \
"t02_S-ISCA_C1_Aedes female-20-07-2017_1_001_456.wav"
wav = wavfile.read(wav_file)
# OPTIONAL: Load the "autoreload" extension so that code can change
%load_ext autoreload
# OPTIONAL: always reload modules so that as you change code ... | _____no_output_____ | FTL | notebooks_tcc/0.3-BrunoGomesCoelho-Show-Normalization-Doesnt-Work.ipynb | BrunoGomesCoelho/mosquito-networking |
Found existing installation: tables 3.4.4
Unistalling tables-3.4.4:
Successfully unistalled tables-3.4.4
Successfully installed tables-3.6.1
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
cd "/content/drive/My Drive/Colab Notebooks/matrix/matrix_two/dw_matrix_car"
df = pd... | _____no_output_____ | Apache-2.0 | day2_visualisation.ipynb | Kamila86/dw_matrix | |
็งปๅใญใใใใฎ็ถๆ
้ท็งป๏ผใใคใบใใ๏ผๅ่ๅทฅๆฅญๅคงๅญฆ ไธ็ฐ ้ไธ(c) 2017 Ryuichi UedaThis software is released under the MIT License, see LICENSE. ใฏใใใซใใฎใณใผใใฏใ็งปๅใญใใใใฎ็งปๅใฎ็ฐกๅใชใขใใซใงใใ | %matplotlib inline
import numpy as np
import math, random
import matplotlib.pyplot as plt # for plotting data | _____no_output_____ | MIT | state_equations/with_noise.ipynb | ryuichiueda/probrobo_practice |
ใฐใใคใใคใใฎใซใผใซใซใใใใฃใฆใญใใใใฎ็งปๅ้ใซ้้ณใๆททๅ
ฅใใฆใฟใใ* ๆไปคๅคใไธญๅฟใซใใฌใฆในๅๅธใซใใใใฃใฆๆฌกใฎใใใซใฐใใคใ * ๅ้ฒใใใจใ * ่ท้ขใ่ท้ขใซๅฏพใใฆ10%ใฎๆจๆบๅๅทฎใงใฐใใคใ * ๅใใๆจๆบๅๅทฎ3[deg]ใงใฐใใคใ * ๅใใๅคใใใจใ * ๅคใใๅใใฎ่งๅบฆใซๅฏพใใฆ10%ใฎๆจๆบๅๅทฎใงใฐใใคใ ็ถๆ
ๆน็จๅผใซๅฏพๅฟใใ้ขๆฐไธใฎไพใซใคใใฆใ้ขๆฐfใๆธใใพใใใใ ```pythonold_x = np.array([0,0,0]) ไปๅใฏไธ่ฆใ ใnumpyใไฝฟ็จu = np.array([0.1,10/180*math.pi]) ๆฏๅ0.1ใ ใ้ฒใใฆ10[deg]ๅใใๅคใใdef... | def f(x_old,u):
# ใใใใซใใใฎใงใใฉใ
pos_x, pos_y, pos_theta = x_old
act_fw, act_rot = u
act_fw = random.gauss(act_fw,act_fw/10)
dir_error = random.gauss(0.0, math.pi / 180.0 * 3.0)
act_rot = random.gauss(act_rot,act_rot/10)
pos_x += act_fw * math.cos(pos_theta + dir_error)
pos_y += a... | _____no_output_____ | MIT | state_equations/with_noise.ipynb | ryuichiueda/probrobo_practice |
ๅฎ่ก๏ผ10ในใใใๅ่ฒธใใฆใฟใพใใใใ | x = np.array([0,0,0])
u = np.array([0.1,10/180*math.pi]) # ๆฏๅ0.1ใ ใ้ฒใใฆ10[deg]ๅใใๅคใใ
print(x)
for i in range(10):
x = f(x,u)
print(x) | [0 0 0]
[ 0.11101391 0.01316534 0.20062588]
[ 0.21667561 0.03772248 0.36210264]
[ 0.30714191 0.06797028 0.53692128]
[ 0.38194774 0.11808143 0.72018441]
[ 0.45733798 0.17144363 0.89551629]
[ 0.5062014 0.23616994 1.07167283]
[ 0.55316268 0.32808052 1.24928927]
[ 0.59204397 0.42298333 1.41256331]
[ 0.6106... | MIT | state_equations/with_noise.ipynb | ryuichiueda/probrobo_practice |
ใใใใใฎใงๆ็ป | x = np.array([0,0,0])
u = np.array([0.1,10/180*math.pi]) # ๆฏๅ0.1ใ ใ้ฒใใฆ10[deg]ๅใใๅคใใ
path = [x]
for i in range(10):
x = f(x,u)
path.append(x)
fig = plt.figure(i,figsize=(8, 8))
sp = fig.add_subplot(111, aspect='equal')
sp.set_xlim(-1.0,1.0)
sp.set_ylim(-0.5,1.5)
xs = [e[0] for e in path]
ys = [e[1] for... | _____no_output_____ | MIT | state_equations/with_noise.ipynb | ryuichiueda/probrobo_practice |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.