How to use test_given method in pytest-bdd

Best Python code snippet using pytest-bdd_python

rand_forr2.py

Source:rand_forr2.py

1#Joseph Day2#3/5/173#Cleaning the data and building a classifier 4import random5import csv6import pandas as pd7import pickle8import numpy as np9import sklearn10from sklearn.metrics import roc_auc_score11from sklearn.metrics import accuracy_score12from sklearn.ensemble import GradientBoostingClassifier as gbc13from sklearn.ensemble import RandomForestClassifier as rfc14df_tour = pd.read_csv('TourneyCompactResults.csv')15df_tour.drop(labels=['Daynum', 'Wscore', 'Lscore', 'Wloc', 'Numot'], inplace=True, axis=1)16df_teams = pd.read_csv('season_averages.csv')17df_seeds = pd.read_csv('TourneySeeds.csv')18def seed_to_int(seed):19    """Get just the digits from the seeding. Return as int"""20    s_int = int(seed[1:3])21    return s_int22df_seeds['n_seed'] = df_seeds.Seed.apply(seed_to_int)23df_seeds.drop(labels=['Seed'], inplace=True, axis=1) 24df_teams['fgp'] = df_teams['fgm']/df_teams['fga']25df_teams['3p'] = df_teams['fgm3']/df_teams['fga3']26df_teams['ftp'] = df_teams['ftm']/df_teams['fta']27#df_teams['seed'] = df_seeds['n_seed']28df_winseeds = df_seeds.rename(columns={'Team':'Wteam', 'n_seed':'win_seed'})29df_lossseeds = df_seeds.rename(columns={'Team':'Lteam', 'n_seed':'loss_seed'})30df_dummy = pd.merge(left=df_tour, right=df_winseeds, how='left', on=['Season', 'Wteam'])31#print(df_dummy.head)32df_concat1 = pd.merge(left=df_dummy, right=df_lossseeds, on=['Season', 'Lteam'])33df_concat1['seed_diff'] = df_concat1.win_seed - df_concat1.loss_seed34#print(df_concat.head)35#df_teams = pd.merge(left=df_teams, right=df_concat, how='left', on=['Season', 'Wteam'])36df_winteams = df_teams.rename(columns={'Team_Id':'Wteam', 'score':'w_score', 'fgp':'w_fgp', '3p':'w_3p',37                'ftp':'w_ftp', 'or':'w_or', 'dr':'w_dr', 'ast':'w_ast', 'to':'w_to', 'stl':'w_stl','blk':'w_blk',38                'pf':'w_pf'})39df_lossteams = df_teams.rename(columns={'Team_Id':'Lteam', 'score':'l_score','fgp':'l_fgp', '3p':'l_3p',40                'ftp':'l_ftp', 'or':'l_or', 'dr':'l_dr', 'ast':'l_ast', 'to':'l_to', 'stl':'l_stl','blk':'l_blk',41                'pf':'l_pf'})42df_dummy = pd.merge(left=df_tour, right=df_winteams, how='left', on=['Season', 'Wteam'])43df_concat = pd.merge(left=df_dummy, right=df_lossteams, on=['Season', 'Lteam'])44df_concat.drop(labels=['Unnamed: 0_x', 'Unnamed: 0_y', 'fgm_x', 'fgm_y', 'fga_x', 'fga_y',45                'fgm3_x', 'fgm3_y', 'fga3_x', 'fga3_y', 'ftm_x', 'ftm_y', 'fta_y', 'fta_x',46                'Team_Name_x', 'Team_Name_y'], inplace=True, axis=1) # This is the string label47df_concat = pd.merge(left=df_concat, right = df_concat1, how='left', on=['Season', 'Wteam', 'Lteam'])48#print(df_concat.head)49df_concat.drop(labels=['win_seed','loss_seed'], inplace=True, axis=1) 50df_concat['pt_diff'] = df_concat.w_score - df_concat.l_score51df_concat['ast_diff'] = df_concat.w_ast - df_concat.l_ast52df_concat['or_diff'] = df_concat.w_or - df_concat.l_or53df_concat['dr_diff'] = df_concat.w_dr - df_concat.l_dr54df_concat['to_diff'] = df_concat.w_to - df_concat.l_to55df_concat['stl_diff'] = df_concat.w_stl - df_concat.l_stl56df_concat['blk_diff'] = df_concat.w_blk - df_concat.l_blk57df_concat['pf_diff'] = df_concat.w_pf - df_concat.l_pf58df_concat['fgp_diff'] = df_concat.w_fgp - df_concat.l_fgp59df_concat['3p_diff'] = df_concat.w_3p - df_concat.l_3p60df_concat['ftp_diff'] = df_concat.w_ftp - df_concat.l_ftp61#df_concat['seed_diff'] = df_concat.w_seed - df_concat.l_seed62df_wins = df_concat[['pt_diff', 'ast_diff', 'or_diff', 'dr_diff', 'to_diff', 'stl_diff', 'blk_diff',63'pf_diff', 'fgp_diff', '3p_diff', 'ftp_diff', 'seed_diff', 'Season', 'Wteam', 'Lteam']]64df_wins['result'] = 165df_losses = -df_concat[['pt_diff', 'ast_diff', 'or_diff', 'dr_diff', 'to_diff', 'stl_diff', 'blk_diff',66'pf_diff', 'fgp_diff', '3p_diff', 'ftp_diff', 'seed_diff', 'Season', 'Wteam', 'Lteam']]67df_losses['result'] = 068df_for_predictions = pd.concat((df_wins, df_losses))69columns = df_for_predictions.columns.tolist()70columns = columns[:-4]71given_data = df_for_predictions.as_matrix(columns = columns)72target_data = np.array(df_for_predictions['result'].tolist())73clf = gbc(n_estimators = 10)74clf = clf.fit(given_data, target_data)75#76df_sample_sub = pd.read_csv('sample_submission.csv')77n_test_games = len(df_sample_sub)78def get_year_t1_t2(id):79    """Return a tuple with ints `year`, `team1` and `team2`."""80    return (int(x) for x in id.split('_'))81X_test = np.zeros(shape=(n_test_games, 12))82for ii, row in df_sample_sub.iterrows():83    year, t1, t2 = get_year_t1_t2(row.id)84    # There absolutely must be a better way of doing this!85    t1_score = df_teams[(df_teams.Team_Id == t1) & (df_teams.Season == year)][['score','ast','or','dr','to','stl','blk','pf','fgp','3p','ftp']].values[0]86    t2_score = df_teams[(df_teams.Team_Id == t2) & (df_teams.Season == year)][['score','ast','or','dr','to','stl','blk','pf','fgp','3p','ftp']].values[0]87    t1_seed = df_seeds[(df_seeds.Team == t1) & (df_seeds.Season == year)].n_seed.values[0]88    t2_seed = df_seeds[(df_seeds.Team == t2) & (df_seeds.Season == year)].n_seed.values[0]89    for i in range(0,11):90        #diff_score.append(t1_score[i] - t2_score[i])91        X_test[ii, i] = t1_score[i] - t2_score[i]92    X_test[ii,11]= t1_seed - t2_seed93preds = clf.predict_proba(X_test)94clipped_preds = np.clip(preds, 0.05, 0.95)95df_sample_sub.pred = preds96df_sample_sub.head()97df_sample_sub.to_csv('rand_forr5.csv', index=False)98def calc_errors(forest, test_file):99    '''Calculates typeI and typeII errors,100    useful in ensuring not too many of either101    '''102    clf = forest103    test_given, test_target = process(test_file, 'responded')104    predictions = clf.predict(test_given)105    typeI = 0106    typeII = 0107    total = 0108    for i in range(1, len(test_target)):109        if predictions[i] == test_target[i]:110            pass111        else:112            if predictions[i] == 'yes':113                typeI += 1114            else: 115                typeII += 1116        total+=1117    typeI = typeI/total118    typeII = typeII/total119    return typeI,typeII120def roc_auc(forest, test_file):121    '''Returns area under ROC curve for 122    classifier provided / built from test data provided123    '''124    clf = forest125    test_given, test_target = process(test_file, 'responded')126    predictions = clf.predict_proba(test_given)127    test_target = pd.DataFrame(test_target)128    test_target.replace(['yes', 'no'], [1,0], inplace=True)129    test_target = np.array(test_target[0].tolist())130    predictions = predictions[:,1]131    return roc_auc_score(test_target, predictions)132def score(forest, test_file):133    '''Function takes in a forest/gbc and a test file134    and returns various metrics for evaluation135    '''136    roc_score = roc_auc(forest, test_file)137    test_given, test_target = process(test_file, 'responded')138    predictions = forest.predict(test_given)139    140    typeI, typeII = calc_errors(forest, test_file)141    accuracy = accuracy_score(test_target, predictions)142    143    print("Roc_Auc_Score is ", roc_score)144    print("Type I Error Rate is ", typeI)145    print("Type II Error Rate is ", typeII)...

analyze.py

Source:analyze.py

1import numpy as np2ratio = 0.9; train_ratio = int(ratio*50)3feature_name = ['annual police funding in $/resident', '% of people 25 years+ with 4 yrs. of high school', '% of 16 to 19 year-olds not in highschool and not highschool graduates', '% of 18 to 24 year-olds in college', '% of people 25 years+ with at least 4 years of college']4train_x = []; test_x = []; train_total = []; test_total = []; train_violent = []; test_violent = []5with open('data.txt', 'r') as to_read:6	did_read = to_read.readlines()7	for line in did_read[:int(ratio*50)]:8		train_x.append(map(lambda x: np.float64(x), line.split()[2:]))9		train_total.append(np.float64(line.split()[0]))10		train_violent.append(np.float64(line.split()[1]))11	for line in did_read[int(ratio*50):]:12		test_x.append(map(lambda x: np.float64(x), line.split()[2:]))13		test_total.append(np.float64(line.split()[0]))14		test_violent.append(np.float64(line.split()[1]))15#print("test_total:"); print(test_total); print(type(test_total))16train_x = np.matrix([d for d in train_x], dtype = float)17test_x = np.matrix([d for d in test_x], dtype = float)18train_total = np.matrix([d for d in train_total], dtype = float)19train_violent = np.matrix([d for d in train_violent], dtype = float)20test_total = np.matrix([d for d in test_total], dtype = float)21test_violent = np.matrix([d for d in test_violent], dtype = float)22#train_x = np.matrix(train_x)23#print(train_x_temp == train_x)24#test_x = np.matrix(test_x)25#train_total = np.matrix(train_total)26#test_total = np.matrix(test_total)27#train_violent = np.matrix(train_violent)28#test_violent = np.matrix(test_violent)29#print("test_total:"); print(test_total); print(type(test_total))30train_x -= np.mean(train_x, 0)31train_x /= np.sqrt(np.sum(np.square(train_x), 0))32test_x -= np.mean(test_x, 0)33test_x /= np.sqrt(np.sum(np.square(test_x), 0))34train_total -= np.mean(train_total)35train_violent -= np.mean(train_violent)36test_total -= np.mean(test_total)37test_violent -= np.mean(test_violent)38#-------------------------------------------------------------------------------------------------------------39def calc_beta(test_given, train_given):40	beta = np.zeros((1000, 5, 1), dtype = float)41	#y = np.zeros((1000, len(test_given), 1), dtype = float)42	y = []43	mse = np.zeros((1000, 1), dtype = float)44	lambdas = np.logspace(-10, 1, 1000)45	#print('train length = {}, test length = {}'.format(len(train_x), len(test_x)))46	for i in range(len(lambdas)):47		beta[i] = (train_x.T*train_x + lambdas[i]*np.identity(5)).I*train_x.T*train_given.T48		#print('beta[i] = ', beta[i].shape, type(beta[i]))49		#print('test_given.shape = ',test_given.shape, type(test_given))50		#print('test_x.shape = ',test_x.shape, type(test_x))51		#y[i] = test_given*beta[i]52		#y[i] = test_x*beta[i]53		to_app = []54		for j in range(5):55			tmp = 056			for k in range(5):57				tmp += (float(test_x.item(j,k)) * float(beta[i][k]))58			to_app.append(tmp)59		y.append(to_app)60		#to_app = map(lambda x:[x], to_app)61		#print('to_app:', to_app, type(to_app))62		#y[i] = np.matrix(to_app)63		#print(y[i])	64		#print('np.square(y[i].shape: '),;print(np.square(y[i]).shape)65		#print('test_given.T.shape'),; print(test_given.T.shape)66		acc = []67		for j in range(5):68			acc.append(y[i][j] - test_given.item(j))69		mse[i]= sum(np.square(acc))70		#mse[i] = sum(np.square(y[i] - test_given.T))/len(test_given)71	idx = np.argmin(mse)72	#print(mse)73	#print 'mse: ',; print(mse[idx]); print 'lambda',; print(lambdas[idx])74	return beta[idx]75#-------------------------------------------------------------------------------------------------------------76if __name__ == "__main__":77	beta = calc_beta(test_total, train_total)78	#print('test_x: '); print(test_x); print('')79	#print('beta: '); print(beta); print('')80	#print('test_total: '); print(test_total); print('')81	err = 0; abs_err = 082	print('Total overall reported crime rate per 1 million residents predictions:')83	for i in range(len(np.array(test_total)[0])):84		predicted = 085		for j in range(5):86			predicted += (beta[j] * test_x.item(i,j))87		err += (np.array(test_total)[0][i] - predicted[0])88		abs_err += abs(np.array(test_total)[0][i] - predicted[0])89		print('actual value: {}, predicted value: {}'.format(np.array(test_total)[0][i], predicted[0]))90	print('Total signed error: {}, absolute error per prediction: {}'.format(err, abs_err/5))91	print 'Prediction coefficients: '92	for i in range(5):93		print('{}) {}: {}'.format(i+1, feature_name[i], beta[i][0]))94	print('----------------------------------------------------------------------------------------------------')	95	beta = calc_beta(test_violent, train_violent)96	err = 0; abs_err = 097	print('Reported violent crime rate per 100,000 residents predictions:')98	for i in range(len(np.array(test_total)[0])):99		predicted = 0100		for j in range(5):101			predicted += (beta[j] * test_x.item(i,j))102		err += (np.array(test_violent)[0][i] - predicted[0])103		abs_err += abs(np.array(test_violent)[0][i] - predicted[0])104		print('actual value: {}, predicted value: {}'.format(np.array(test_violent)[0][i], predicted[0]))105	print('Total signed error: {}, absolute error per prediction: {}'.format(err, abs_err/5))106	print 'Prediction coefficients: '107	for i in range(5):...

main.py

Source:main.py

1import argparse2import reader3from keras.models import Sequential4from keras.layers.core import Dense, Activation, Dropout5from keras.layers.recurrent import LSTM6from keras.utils.data_utils import get_file7from keras.layers import Embedding8import json9import numpy as np10import random11import sys12import os13import shutil14import requests15parser = argparse.ArgumentParser()16parser.add_argument("--test", help="path of test.txt file")17args = parser.parse_args()18f=open('word-ix.txt', 'r')19word_to_ix=json.loads(f.read())20f.close()21model = Sequential()22model.add(Embedding(10000, output_dim=64))23model.add(LSTM(100, return_sequences=False))24model.add(Dropout(0.2))25model.add(Dense(10000))26model.add(Activation('softmax'))27model.compile(loss='categorical_crossentropy', optimizer='rmsprop')28if args.test:29    test_given=reader._file_to_word_ids(args.test, word_to_ix)30    maxlen = 1531    step = 532    sentences=[]33    next_words=[]34    for i in range(0,len(test_given)-maxlen, 1):35        sentences2 = (test_given[i: i + maxlen])36        sentences.append(sentences2)37        next_words.append((test_given[i + maxlen]))38    X_test = np.zeros((len(sentences), maxlen))39    y_test = np.zeros((len(sentences)), dtype=np.int32)40    for i, sentence in enumerate(sentences):41        for t, word in enumerate(sentence):42            X_test[i, t] = word43        y_test[i]=next_words[i]44    if os.path.isfile('weights'):45        model.load_weights('weights')46    else:47        print ("Downloading Model ......")48        url = 'https://raw.githubusercontent.com/shivam207/deep_learning/master/prog_asg3/weights/model_cnn.h5'49        response = requests.get(url, stream=True)50        with open('model_cnn.h5', 'w') as out_file:51            shutil.copyfileobj(response.raw, out_file)52    del response53    preds = model.predict(X_test, verbose=0)54    loss = 055    for i, pred in enumerate(preds):56        loss += -pred[y_test[i]]*np.log(pred[y_test[i]])57    print loss58else:59    train_data, valid_data, test_data, vocabulary, word_to_ix = reader.ptb_raw_data("data")60    words = set(train_data)61    words.union(valid_data)62    words.union(test_data)63    maxlen = 1564    step = 565    print("maxlen:",maxlen,"step:", step)66    sentences = []67    next_words = []68    sentences2=[]69    for i in range(0,len(train_data)-maxlen, step):70        sentences2 = (train_data[i: i + maxlen])71        sentences.append(sentences2)72        next_words.append((train_data[i + maxlen]))73    print('nb sequences(length of sentences):', len(sentences))74    print("length of next_word",len(next_words))75    print('Vectorization...')76    X = np.zeros((len(sentences), maxlen))77    y = np.zeros((len(sentences), len(words)), dtype=np.bool)78    for i, sentence in enumerate(sentences):79        for t, word in enumerate(sentence):80            #print(i,t,word)81            X[i, t] = word82        y[i, next_words[i]] = 183    for iteration in range(1, 300):84        print()85        print('-' * 50)86        print('Iteration', iteration)87        model.fit(X, y, batch_size=200 ,nb_epoch=2)...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.