Join the 1-day Testing & QA Summit featuring 15+ Expert Speakers.Register for FREE! Join TestMu Conference

How to use test_positive method in hypothesis

Best Python code snippet using hypothesis

data_split.py

Source:data_split.py

1import os2import pandas as pd3import numpy as np4import argparse5import random67parser = argparse.ArgumentParser(description='data_split')8parser.add_argument('--data_path', default='./Amazon-office-raw/ratings.csv', type=str,9                    help='datapath')10parser.add_argument('--save_path', default='./Amazon-office-raw', type=str,11                    help='savepath')12parser.add_argument('--negative_sampling', default=True, type=bool,13                    help='test negative sampling')14args = parser.parse_args()15random.seed(1)1617ratings=pd.read_csv(args.data_path)18type=['ratio-split','leave-one-out']19# indexing user20user_id=ratings[['userid']].drop_duplicates().reset_index(drop=True)21user_id['useridx']=pd.DataFrame(np.arange(len(user_id)))22userdict=dict(zip(user_id['userid'],user_id['useridx']))23ratings=ratings.replace({'userid':userdict})2425#indexing item26item_id=ratings[['itemid']].drop_duplicates().reset_index(drop=True)27item_id['itemidx']=pd.DataFrame(np.arange(len(item_id)))28itemdict=dict(zip(item_id['itemid'],item_id['itemidx']))29ratings=ratings.replace({'itemid':itemdict})3031# group items according to user index32items_set=set(item_id['itemidx'])33#ratings['rate'][ratings['rate']>0]=1.034rating1=ratings.groupby('userid')['itemid'].apply(list).reset_index()35rating2=ratings.groupby('userid')['timestamp'].apply(list).reset_index()3637# split38for split_type in type:39    ratings = pd.concat([rating1[['userid', 'itemid']], rating2['timestamp']], axis=1)40    if split_type=='leave-one-out':41        ratings['test_positive_index']=ratings['timestamp'].apply(lambda x: np.array(x).argmax())42        ratings['test_positive']=ratings.apply(lambda x: x['itemid'][x['test_positive_index']], axis=1)43        ratings['negative']=ratings['itemid'].apply(lambda x: list(items_set - set(x)))44        ratings['test_negative']=ratings['negative'].apply(lambda x: random.sample(x, 99))45        ratings['train_negative']=ratings.apply(lambda x: list(items_set - set(x['itemid']) - set(x['test_negative'])), axis=1)46        ratings.apply(lambda x: x['itemid'].remove(x['test_positive']), axis=1)474849    elif split_type=='ratio-split':50        ratings['test_positive']=ratings['itemid'].apply(lambda x: random.sample(x, round(len(x)*0.2)))51        ratings['itemid']=ratings.apply(lambda x: list(set(x['itemid'])-set(x['test_positive'])), axis=1)52        ratings['train_negative']=ratings.apply(lambda x: list(items_set - set(x['itemid'])),axis=1)53        if args.negative_sampling:54            ratings['test_negative']=ratings.apply(lambda x: random.sample(list(items_set - set(x['itemid'])- set(x['test_positive'])),len(x['test_positive']*10))55                                                if len(x['test_positive'])*10 <= len(list(items_set-set(x['itemid'])-set(x['test_positive']))) else list(items_set - set(x['itemid'])- set(x['test_positive'])), axis=1)5657        else :58            ratings['test_negative']=ratings.apply(lambda x: list(items_set - set(x['itemid'])- set(x['test_positive'])), axis=1)59        ratings['train_negative'] = ratings.apply(lambda x: list(set(x['train_negative'])-set(x['test_negative'])), axis=1)6061    ratings.rename(columns = {'itemid':'train_positive'}, inplace = True)62    ratings = ratings[['userid','train_positive','test_positive','train_negative', 'test_negative']].reset_index(drop=True)63    train_positive=ratings.join(ratings['train_positive'].apply(lambda x:pd.Series(x)).stack().reset_index(1,name='train_pos').drop('level_1',axis=1))64    test_positive=ratings.join(ratings['test_positive'].apply(lambda x:pd.Series(x)).stack().reset_index(1,name='test_pos').drop('level_1',axis=1))6566    # save67    if not os.path.exists(os.path.join(args.save_path, split_type)):68        os.makedirs(os.path.join(args.save_path, split_type))69    train_positive[['userid','train_pos']].reset_index(drop=True).to_feather(os.path.join(args.save_path,split_type, 'train_positive.ftr'))70    test_positive[['userid','test_pos']].reset_index(drop=True).to_feather(os.path.join(args.save_path,split_type, 'test_positive.ftr'))71    ratings[['userid','train_negative']].reset_index(drop=True).to_feather(os.path.join(args.save_path,split_type, 'train_negative.ftr'))72    ratings[['userid','test_negative']].reset_index(drop=True).to_feather(os.path.join(args.save_path,split_type, 'test_negative.ftr'))7374if not os.path.exists(os.path.join(args.save_path,'index-info')):75    os.makedirs(os.path.join(args.save_path,'index-info'))76user_id.to_csv(os.path.join(args.save_path,'index-info', 'user_index.csv'),index=False)
...

evaluation_random_classifier.py

Source:evaluation_random_classifier.py

1import json2import subprocess3import sys4import os5import time6from PIL import Image7from numpy import average, dot, linalg8import cv29import skimage.measure10import os11import subprocess12import random13import numpy14# give a list for psc2code video dealing time15# def getlist_psc2code():16#     output = {}17#     rightlist = []18#     faultlist = []19#     database = "data_copy/test"20#     for category in os.listdir(database):21#         # print(category) 22#         if category == "non-screencast":23#             for video in os.listdir(os.path.join(database,category)):24#                 rightlist.append(video)25#                 for img in os.listdir(os.path.join(database,category,video)):26#                     # if "frame" not in img:27#                         # print(img)28#                     output[video] = int(1.6*len(os.listdir(os.path.join(database,category,video))))29#         else:30#             for video in os.listdir(os.path.join(database,category)):31#                 faultlist.append(video)32#                 for img in os.listdir(os.path.join(database,category,video)):33#                     # if "frame" not in img:34#                         # print(img)35#                     output[video] = len(os.listdir(os.path.join(database,category,video)))36#     return output,rightlist,faultlist37# psc2codelist,nonlist,yeslist = getlist_psc2code()38# count = 039# for video in psc2codelist.keys():40#     psc2codelist[video] = int(psc2codelist[video]*1.88)41#     count+= psc2codelist[video]42# print(len(nonlist))43# print(len(yeslist))44# # wrong one: MkJ1jY8ubKs45# # 2.5å46# def random1():47#     count = 048#     test_positive = random.sample(list(psc2codelist), k=13)49#     test_negative = list(set(list(psc2codelist)).difference(set(test_positive)))50#     # print(len(list(psc2codelist)))51#     # print(len(test_positive))52#     # print(len(test_negative))53#     tp = 054#     fp = 055#     fn = 056#     for item in test_positive:57#         if item in yeslist:58#             tp += 159#         if item in nonlist:60#             fp+=161#     for item in test_negative:62#         if item in yeslist:63#             fn+=164#     return tp,fp,fn65# allpre = []66# allrecall = []67# allfi = []68# for i in range(20):69#     tp,fp,fn = random1()70#     pre = float(tp)/(tp+fp)71#     allpre.append(pre)72#     recall = float(tp)/(tp+fn)73#     allrecall.append(recall)74#     allfi.append(2*recall*pre/(recall+pre))75#     # print("%d,%d,%d"%(tp,fp,fn))76# print("the pre is %.3f, the recall is %.3f, the F-1 is %.3f"%(numpy.mean(allpre),numpy.mean(allrecall),numpy.mean(allfi))) 77def random1():78    count = 079    psc2codelist = [i for i in range(22)]80    print(psc2codelist)81    yeslist = [i for i in range(14)]82    nonlist = [i for i in range(14,22)]83    print(yeslist)84    print(nonlist)85    test_positive = random.sample(list(psc2codelist), k=11)86    test_negative = list(set(list(psc2codelist)).difference(set(test_positive)))87    # print(len(list(psc2codelist)))88    # print(len(test_positive))89    # print(len(test_negative))90    tp = 091    fp = 092    fn = 093    for item in test_positive:94        if item in yeslist:95            tp += 196        if item in nonlist:97            fp+=198    for item in test_negative:99        if item in yeslist:100            fn+=1101    return tp,fp,fn102allpre = []103allrecall = []104allfi = []105for i in range(20):106    tp,fp,fn = random1()107    pre = float(tp)/(tp+fp)108    allpre.append(pre)109    recall = float(tp)/(tp+fn)110    allrecall.append(recall)111    allfi.append(2*recall*pre/(recall+pre))112    # print("%d,%d,%d"%(tp,fp,fn))...

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.