KaneAI - World’s First E2E Software Testing Agent.Signup for Private Beta Join TestMu Conference

How to use test_transformations method in Slash

Best Python code snippet using slash

library.py

Source:library.py

1import pickle 2import os3import math4import collections5import random6class TopicSeries(list):7    """8    Derived "list" class, with ability to return9    all subsets of a particular length10    """11    def get_subseries(self,length):12        """ Generator that returns all sub-lists of self 13        that are length "length"14        """15        index = 016        while index + length <= len(self):17            yield self[index:index+length]18            index += 119class Library(object):20    def __init__(self, **kwargs):21        """22        set up internal lists, get parameter values, 23        and set up transformation functions24        """25        self.trends = []26        self.non_trends = []27        28        self.config = {}29        # default values30        self.config["reference_length"] = 21031        self.config["n_smooth"] = 8032        self.config["alpha"] = 1.233        # add values passed into ctor34        self.config.update(kwargs["config"])35        # Transformation functions are defined globally and added here to list.36        37        # transformations to be run on reference series38        self.transformations = []39        self.transformations.append(add_one)40        self.transformations.append(unit_normalization)41        self.transformations.append(logarithmic_scaling)42        self.transformations.append(smoothing)43        self.transformations.append(sizing)44        45        # transformations to be run on test series 46        self.test_transformations = []47        self.test_transformations.append(add_one)48        self.test_transformations.append(unit_normalization)49        self.test_transformations.append(logarithmic_scaling)50        self.test_transformations.append(smoothing)51       52    def add_reference_series(self,series,is_trend=True):53        """54        add a reference time series to the internal lists,55        after transforming it 56        """57        self.config["is_trend"] = is_trend58        series = self.transform_input(series,is_test_series=False) 59        if is_trend:60            self.trends.append( TopicSeries(series) )61        else:62            self.non_trends.append( TopicSeries(series) )63    def transform_input(self,series,is_test_series,config=None):64        """65        Run series sequentially through the functions 66        in the transformations list67        """68        transformations = self.transformations69        if is_test_series:70            transformations = self.test_transformations71        72        for transformation in transformations:73            if config is not None:74                series = transformation(series,config) 75            else:76                series = transformation(series,self.config)77        return series78    def combine(self, lib):79        """80        Manage all attributes of class that are important for combinations. 81        Take care not to allow duplicates.82        """83        if lib.trends != []:84            assert self.trends == []85            self.trends = lib.trends86        if lib.non_trends != []: 87            assert self.non_trends == []88            self.non_trends = lib.non_trends89def add_one(series, config):90    """ Add a count of 1 to every count in the series """91    return [ ct+1 for ct in series ]92def unit_normalization(series, config):93    """ Do unit normalization based on "reference_length" number of bins94    at the end of the series"""95    reference_length = int(config["reference_length"])96    SMALL_NUMBER = 0.0000197    offset = int(config["baseline_offset"])98    lower_idx = -(int(config["reference_length"]) + offset)99    upper_idx = -offset100    total = sum(series[lower_idx:upper_idx])/float(reference_length)101    if total == 0:102        total = SMALL_NUMBER103    return [float(pt)/total for pt in series]104def spike_normalization(series, config):105    alpha = float(config["alpha"])106    new_series = []107    prev_pt = 0108    for pt in series: 109        if pt == 0:110            new_pt = 0111        else:112            new_pt = math.pow(abs(pt - prev_pt), alpha)113        new_series.append(new_pt)114        prev_pt = pt115    return new_series116def smoothing(series,config): 117    n_smooth = int(config["n_smooth"])118    queue = collections.deque()119    new_series = []120    for pt in series:121        queue.append(pt)122        new_series.append( float(sum(queue))/len(queue) )123        if len(queue) >= n_smooth:124            queue.popleft()125    return new_series126def slow_smoothing(series,config): 127    n_smooth = int(config["n_smooth"])128    queue = []129    new_series = []130    for pt in series:131        queue.append(pt)132        new_series.append( float(sum(queue))/len(queue) )133        if len(queue) >= n_smooth:134            del queue[0]135    return new_series136def index_smoothing(series,config): 137    n_smooth = int(config["n_smooth"])138    new_series = []139    idx = 1140    while idx < len(series):141        lower_idx = max(0,idx-n_smooth)142        sub_series = series[lower_idx:idx]143        new_series.append( float(sum(sub_series))/len(sub_series) )144        idx+=1145    return new_series146def logarithmic_scaling(series, config): 147    new_series = []148    149    for pt in series:150        if pt <= 0:151            pt = 0.00001152        new_series.append(math.log10(pt))153    return new_series154def sizing(series, config): 155    new_series = series[-int(config["reference_length"]):]156    return new_series157def save_library(library, file_name):158    pickle.dump(library,open(file_name,"w"))159def load_library(file_name):160    try:161        return pickle.load(open(file_name)) 162    except EOFError:163        return Library()164def merge_library(library, file_name): 165    """166    if file exists, get Library object from it,167    and combine with library passed to function168    """169    if os.path.exists( os.path.join(os.getcwd(),file_name) ):170        lib_from_file = load_library(file_name)171        library.combine(lib_from_file)172    return library173if __name__ == "__main__":174    import sys175    import argparse176    parser = argparse.ArgumentParser() 177    parser.add_argument("-t",dest="is_trend",default=False,action="store_true")178    parser.add_argument("-f",dest="lib_file_name",default="library.pkl")179    args = parser.parse_args()180    series = []181    for ct in sys.stdin: 182        series.append(ct)183    lib = Library()184    lib.add_reference_series(series,trend = args.is_trend)185    merge_library(lib,args.lib_file_name)...

torchvision_dataset.py

Source:torchvision_dataset.py

1import torch2from torchvision import datasets, transforms, models3from collections.abc import Iterable4# this shit seems to be required to download5from six.moves import urllib6opener = urllib.request.build_opener()7opener.addheaders = [('User-agent', 'Mozilla/5.0')]8urllib.request.install_opener(opener)9def get_datasets(name, batch_size_train=256, batch_size_test=1024,10                 num_workers=2, pin_memory=True, transformation_kwargs=None):11    # TODO: validation12    dataset = getattr(datasets, name)13    if transformation_kwargs is None:14        transformation_kwargs = {}15    train_transform, test_transform = get_transformations(**transformation_kwargs)16    train_dataset = dataset(f'{name.lower()}_data', train=True, download=True,17                transform=train_transform)18    test_dataset = dataset(f'{name.lower()}_data', train=False, download=False,19                transform=test_transform)20    return train_dataset, test_dataset21def get_transformations(flip=False, crop=False, crop_size=32, crop_padding=4, normalize=None):22    train_transformations = []23    test_transformations = []24    if flip:25        train_transformations.append(transforms.RandomHorizontalFlip())26    if crop:27        train_transformations.append(transforms.RandomCrop(crop_size, padding=crop_padding))28    # to tensor29    train_transformations.append(transforms.ToTensor())30    test_transformations.append(transforms.ToTensor())31    if normalize == 'cifar':32        train_transformations.append(transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)))33        test_transformations.append(transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)))34    elif isinstance(normalize, Iterable):35        train_transformations.append(transforms.Normalize(normalize[0], normalize[1]))36        test_transformations.append(transforms.Normalize(normalize[0], normalize[1]))37    train_transform = transforms.Compose(train_transformations)38    test_transform = transforms.Compose(test_transformations)...

cases_transformations_called.py

Source:cases_transformations_called.py

1from pytest_cases import case2@case(id="squash_columns called")3def case_squash():4    test_transformations = {5        "squash_columns": [6            {7                "original_columns": ["Dep Adrs1", "Dep Adrs2", "Dep Adrs3"],8                "aggregate_col": "address_lines",9            }10        ],11    }12    log_message = "mock squash_columns"13    return (test_transformations, log_message)14@case(id="convert_to_bool called")15def case_bool():16    test_transformations = {17        "convert_to_bool": [18            {19                "original_columns": ["Dep Adrs1", "Dep Adrs2", "Dep Adrs3"],20                "aggregate_col": "address_lines",21            }22        ],23    }24    log_message = "mock convert_to_bool"25    return (test_transformations, log_message)26@case(id="unique_number called")27def case_unique():28    test_transformations = {29        "unique_number": [30            {31                "original_columns": ["Dep Adrs1", "Dep Adrs2", "Dep Adrs3"],32                "aggregate_col": "address_lines",33            }34        ],35    }36    log_message = "mock unique_number"...

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.