How to use test_z1 method in hypothesis

Best Python code snippet using hypothesis

experiment.py

Source:experiment.py

1import os2import pickle5 as pickle3import pandas as pd4import numpy as np5from sklearn.linear_model import LogisticRegression6from sklearn.ensemble import RandomForestClassifier7def preprocess(tp, dir_path):8    """9    :param dir_path: ~/VFAE10    :param tp: 0 -> german, 1 -> adult, 2 -> health11    :return: scores of logistic, random forest, random choice & accuracy on y12    """13    if tp == 0:14        data = 'german'15        sensitive = 'sex'16        target_dir = dir_path + '/dataset/german/'17        with open(os.path.join(target_dir, 'german_train.pkl'), 'rb') as f:18            train = pickle.load(f)19        with open(os.path.join(target_dir, 'german_test.pkl'), 'rb') as f:20            test = pickle.load(f)21    elif tp == 1:22        data = 'adult'23        sensitive = 'sex'24        target_dir = dir_path + '/dataset/adult/'25        with open(os.path.join(target_dir, 'adult_train.pkl'), 'rb') as f:26            train = pickle.load(f)27        with open(os.path.join(target_dir, 'adult_test.pkl'), 'rb') as f:28            test = pickle.load(f)29    else:30        data = 'health'31        target_dir = dir_path + '/dataset/health/'32        sensitive = 'age'33        with open(os.path.join(target_dir, 'health_train.pkl'), 'rb') as f:34            train = pickle.load(f)35        with open(os.path.join(target_dir, 'health_test.pkl'), 'rb') as f:36            test = pickle.load(f)37    train_z1_vae = np.load(os.path.join(dir_path + '/exp_%s/' % data, "%s_train_z1_vae.npy" % data))38    test_z1_vae = np.load(os.path.join(dir_path + '/exp_%s/' % data, "%s_test_z1_vae.npy" % data))39    train_z1_vfae = np.load(os.path.join(dir_path + '/exp_%s/' % data, "%s_train_z1_vfae.npy" % data))40    test_z1_vfae = np.load(os.path.join(dir_path + '/exp_%s/' % data, "%s_test_z1_vfae.npy" % data))41    train_s = train.pop(sensitive).astype('category')42    train_y = train.pop('label').astype('category')43    train_x = train.astype('category')44    test_s = test.pop(sensitive).astype('category')45    test_y = test.pop('label').astype('category')46    test_x = test.astype('category')47    train_z1 = [train_x, train_z1_vae, train_z1_vfae]48    test_z1 = [test_x, test_z1_vae, test_z1_vfae]49    return [train_z1, test_z1], [train_s, test_s], [train_y, test_y]50def calculate(z, s, y):51    [train_z1, test_z1] = z52    [train_s, test_s] = s53    [train_y, test_y] = y54    log_scores = []55    rf_scores = []56    log_y_scores = []57    disc_scores = []58    disc_prob_scores = []59    for i in range(3):60        # logistic regression & random forest61        log = LogisticRegression(random_state=1).fit(train_z1[i], train_s)62        log_scores.append(log.score(test_z1[i], test_s))63        rf = RandomForestClassifier(random_state=1).fit(train_z1[i], train_s)64        rf_scores.append(rf.score(test_z1[i], test_s))65        log_y = LogisticRegression(random_state=1).fit(train_z1[i], train_y)66        log_y_scores.append(log_y.score(test_z1[i], test_y))67        # discrimination on s68        pred = log_y.predict(test_z1[i])69        s_zero = test_s == 070        s_one = test_s == 171        disc_scores.append(np.abs(((pred[s_zero] == test_y[s_zero]).sum()) / s_zero.sum() -72                                  ((pred[s_one] == test_y[s_one]).sum()) / s_one.sum()))73        # discrimination prob on s74        pred_probs = log_y.predict_proba(test_z1[i])75        pred_prob = pd.Series(np.apply_along_axis(lambda x: x[0] if x[0] > x[1] else x[1], 1, pred_probs), dtype=float)76        disc_prob_scores.append(np.abs((pred_prob[s_zero].sum()) / s_zero.sum() -77                                       (pred_prob[s_one].sum()) / s_one.sum()))...

main.py

Source:main.py

1#!/usr/bin/env python22# -*- coding: utf-8 -*-3"""4Created on Sun Apr  8 21:49:30 20185@author: abinaya6"""7import numpy as np8import matplotlib.pyplot as plt9import math10from scipy.stats import norm, uniform11def function_f(x):12    z = (3 - 0.8) * (1/ (1 + (np.sinh(2*x) * np.log(x))))13    return z14 15def function_p(x):16    z = uniform.pdf(x, loc=0.8, scale=3)17    return z18def function_g(x, mu, std):19    z = norm.pdf(x, loc=mu, scale=std)20    return z21print "MONTE CARLO ------------------------"22max_iterations = 5023monte_carlo_estimates = []24for i in range(0,max_iterations):    25    test_x1 = np.random.uniform(0.8,3,size=1000)26    #test_z1 = map(lambda t: function_f(t),test_x1)27    test_z1 = (3 - 0.8) * (1/ (1 + (np.sinh(2*test_x1) * np.log(test_x1))))28    monte_carlo_estimates.append(np.mean(test_z1))29    30print "Mean Integration Value- Simple Monte Carlo: ", np.mean(monte_carlo_estimates)31print "Variance Integration Value- Simple Monte Carlo: ", np.var(monte_carlo_estimates)32'''33test_x1 = np.linspace(0.8,3,1000)34test_z1 = (3 - 0.8) * (1/ (1 + (np.sinh(2*test_x1) * np.log(test_x1))))35plt.figure()36plt.plot(test_x1,test_z1)37plt.xlabel('Samples')38plt.ylabel('Function Value')39plt.title('Integration Function Plot - Stratification')40'''41print "\nMONTE CARLO USING STRATIFICATION ------------------------"42n=100043n1=70044n2=30045value1 = []46value2 = []47for i in range (0,50):48    X1_1 = np.random.uniform(0.8, 1.24, n1) 49    Fnc1_1 = (1.24-0.8) * pow((1 + (np.sinh(2*X1_1)*np.log(X1_1))),-1) 50    value1_1 = (np.sum(Fnc1_1))/ n151    52    X1_2 = np.random.uniform(1.24, 3, n2) 53    Fnc1_2 = (3-1.24)*pow((1 + (np.sinh(2*X1_2)*np.log(X1_2))),-1) 54    value1_2 = (np.sum(Fnc1_2))/ n255    56    value2.append(value1_1 + value1_2)57variance_stratified_Fnc1 = np.var(value2)58Mean_stratified_Fnc1 = np.mean(value2)59print  "Mean Integration Value- Stratification: ", Mean_stratified_Fnc160print  "Variance Integration Value- Stratification: ", variance_stratified_Fnc161'''62plt.scatter(X1_1,Fnc1_1, color='r')63plt.scatter(X1_2,Fnc1_2, color='g')64'''65print "\nMONTE CARLO USING IMPORTANCE SAMPLING ------------------------"66importance_sampling_estimates = []67mu = 0.568std = 0.769for i in range(0,max_iterations):70    test_x3 = np.random.normal(loc=mu, scale=std, size=1000)71    72    fx3 =  (3 - 0.8) * (1/ (1 + (np.sinh(2*test_x3) * np.log(test_x3))))73    px3 = uniform.pdf(test_x3, loc=0.8, scale=3)74    gx3 = norm.pdf(test_x3, loc=mu, scale=std)75    76    z3 = (np.array(fx3)) * (np.array(px3) / np.array(gx3))77    z3 = z3[~np.isnan(z3)]78    importance_sampling_estimates.append(np.mean(z3))79    80print  "Mean Integration Value- Importance sampling: ", np.mean(importance_sampling_estimates)...

test_given_reuse.py

Source:test_given_reuse.py

...23    pass24given_named_booleans = given(z=st.text())25def test_fail_independently():26    @given_named_booleans27    def test_z1(z):28        assert False29    @given_named_booleans30    def test_z2(z):31        pass32    with pytest.raises(AssertionError):33        test_z1()...

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.