How to use compose_feature method in lettuce-tools

Best Python code snippet using lettuce-tools_python

mvFeatureExtractor.py

Source:mvFeatureExtractor.py

...22    elif (chart['chart_type'] == 'scatter'):23        return [0,1,0,0,0]24    elif (chart['chart_type'] == 'area'):25        return [1,0,0,0,0]26def compose_feature(chart, mv_charts):27    def if_decompose(c1, c2):28        return all(i in c2 for i in c1)  ## c1 in c229    mv_charts = [c for c in  mv_charts if chart['indices'] != c['indices'] and chart['chart_type'] != c['chart_type']]30    return sum(1 for c in mv_charts if if_decompose(chart, c))31def complementary_feature(chart, mv_charts):32    def if_complementary(pair):33        return len(set(itertools.chain(*pair))) == sum([len(x) for x in pair])34    mv_charts = [c for c in mv_charts if chart['indices'] != c['indices'] and chart['chart_type'] != c['chart_type']]35    combinations = [(chart, c) for c in mv_charts]36    return sum(1 for x in combinations if if_complementary(x))37def chart_to_feature(chart, mv_charts, all_charts_with_normed_score):38    chart_type_n_columns_feature = chart_type_feature(chart)39    score = [x for x in all_charts_with_normed_score if x['indices'] == chart['indices'] and x['chart_type'] == chart['chart_type']][0]['final_score']40    41    # # print(len(chart['indices']), chart_type_feature(chart), compose_feature(chart,mv_charts), 42    # #         complementary_feature(chart,mv_charts), score)43    # print(chart)44    return [len(chart['indices']), *chart_type_feature(chart), compose_feature(chart,mv_charts), 45            complementary_feature(chart,mv_charts), score]46def charts_to_features_dl(mv_charts, all_charts_with_normed_score, seq_length = False):47    features = []48    for c in mv_charts:49        features.append(chart_to_feature(c, mv_charts, all_charts_with_normed_score))50    if seq_length != False:51        ## padding zero to match seq_length52        padding_zero = np.zeros((seq_length - len(features), len(features[0]))).tolist()53        features.extend(padding_zero)54    return features55def get_chart_lists(raw_lists):56    def get_chart(raw_chart):57        if raw_chart['markEncoding'] == 'arc':58            chart_type = 'pie'...

resampling.py

Source:resampling.py

1import numpy as np2import pandas as pd3from data_process.ProcessedData import ProcessedData4class ResamplingData(ProcessedData):5    def __init__(self, raw_data):6        super().__init__(raw_data)7        self.rest_columns = raw_data.rest_columns8    def process(self):9        equal_zero_index = (self.label_df != 1).values10        equal_one_index = ~equal_zero_index11        pass_feature = np.array(self.feature_df[equal_zero_index])12        fail_feature = np.array(self.feature_df[equal_one_index])13        diff_num = len(pass_feature) - len(fail_feature)14        if diff_num < 1 or len(fail_feature) <= 0:15            return16        temp_array = np.zeros([diff_num, len(self.feature_df.values[0])])17        for i in range(diff_num):18            temp_array[i] = fail_feature[i % len(fail_feature)]19        features_np = np.array(self.feature_df)20        compose_feature = np.vstack((features_np, temp_array))21        label_np = np.array(self.label_df)22        gen_label = np.ones(diff_num).reshape((-1, 1))23        compose_label = np.vstack((label_np, gen_label))24        self.label_df = pd.DataFrame(compose_label, columns=['error'], dtype=float)25        self.feature_df = pd.DataFrame(compose_feature, columns=self.feature_df.columns, dtype=float)...

CLIP_model.py

Source:CLIP_model.py

1import torch2import clip3import torch.nn as nn4class clip_model(nn.Module):5    def __init__(self, image_dim, text_dim):6        super().__init__()7        self.clip_model, _ = clip.load("ViT-B/32")8        # self.compose_linear = nn.Linear(image_dim+text_dim, image_dim)9    def forward(self, image, text):10        image_feature = self.clip_model.encode_image(image) # batch x 51211        text_feature = self.clip_model.encode_text(clip.tokenize(text).cuda()) # batch x 51212        compose_feature = torch.cat((image_feature, text_feature), dim=-1)13        # compose_feature = self.compose_linear(compose_feature)...

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.