How to use copy_tree method in Airtest

Best Python code snippet using Airtest

syn_2.py

Source:syn_2.py

1from _csv import reader2import matplotlib.pyplot as plt3import numpy as np4import pandas as pd5from numpy import where6from numpy import meshgrid7from numpy import arange8from numpy import hstack91011gini_in={}12info_gain={}13b_scoreG=[]14b_valueG=[]15depth_c=316count_acc=017def split_lr(index, value, dataset):18    left, right = list(), list()19    for row in dataset:20        if row[index] < value:21            left.append(row)22        else:23            right.append(row)24    return left, right252627# Gini_index calculating28def gini_impurity(groups, classes):29    n_instances = float(sum([len(group) for group in groups]))30    gini = 0.031    for group in groups:32        size = float(len(group))33        if size == 0:34            continue35        score = 0.036        for class_val in classes: # calculating the GINI Impurity37            p = [row[-1] for row in group].count(class_val) / size38            score += p * p39        gini += (1.0 - score) * (size / n_instances)40    if len(gini_in) != 0: #calculating the information gain41        gini = gini_in[b_valueG[len(b_valueG)-1]]-gini4243    return gini,score444546# Spliting the dataset47def spliting(dataset):48    class_count_left=049    class_count_right=050    class_values = list(set(row[-1] for row in dataset))51    row_index, row_value, gini_score, split_groups,score_t = 999, 999, 999, None,None52    for index in range(len(dataset[0]) - 1):53        for row in dataset:54            groups =split_lr(index, row[index], dataset)55            gini , score= gini_impurity(groups, class_values)56            if len(gini_in)==0:57                if gini < gini_score: # for the root node58                    row_index, row_value, gini_score, split_groups,score_t  = index, row[index], gini, groups,score59            else:60                gini_score=0 #getting the values of information gain for each split and selecting the once with the higher value61                if gini>gini_score:62                    row_index, row_value, gini_score, split_groups, score_t = index, row[index], gini, groups, score6364    gini_in[row_value]=gini_score65    b_scoreG.append(score_t)66    b_valueG.append(row_value)67    # if len(gini_in)==0:68    #     info_gain[]69    # print("gini",gini_in[b_scoreG[0]])70    left,right=split_groups7172    # outcomes = [row[-1] for row in left]73    # class_count_left=outcomes.count()74    # outcomes = [row[-1] for row in right]75    # class_count_right = outcomes.count()7677    return {'index': row_index, 'value': row_value, 'groups': split_groups,'count_left':0,'count_right':0}787980#terminal node81def to_terminal(group):82    outcomes = [row[-1] for row in group]83    count=084    for i in outcomes:85        if i==max(set(outcomes), key=outcomes.count):86            count+=187    return max(set(outcomes), key=outcomes.count),count888990# Create child splits for a node or make terminal, pruning based on the max depth and minimum size is done in this function91def split(node, max_depth, min_size, depth):92    left, right = node['groups']93    del (node['groups'])94    # check for a no split95    if not left or not right:96        val_left,counter=to_terminal(left + right)97        node['left'],node['count_left'] = node['right'],node['count_right'] = val_left,counter98        return99    # check for max depth100    if depth >= max_depth:101        val_left, counter_left = to_terminal(right)102        val_right,counter_right=to_terminal(right)103        node['left'],node['count_left'], node['right'],node['count_right'] = val_left, counter_left,val_right,counter_right104        return105    # process left child106    if len(left) <= min_size:107        val_left, counter_left = to_terminal(right)108        node['left'],node['count_left'] = val_left, counter_left109    else:110        node['left'] = spliting(left)111        split(node['left'], max_depth, min_size, depth + 1)112    # process right child113    if len(right) <= min_size:114        val_right, counter_right = to_terminal(right)115        node['right'],node['counter_right'] =  val_right, counter_right116    else:117        node['right'] = spliting(right)118        split(node['right'], max_depth, min_size, depth + 1)119120121# Build a decision tree122def build_tree(train, max_depth, min_size):123    root = spliting(train)124    split(root, max_depth, min_size, 1)125    return root126127def predict(node, row):128    test = float(row[node['index']])129    if  float(row[node['index']]) < float(node['value']):130        if isinstance(node['left'], dict):131            return predict(node['left'], row)132        else:133            return node['left']134    else:135        if isinstance(node['right'], dict):136            return predict(node['right'], row)137        else:138            return node['right']139140141# Print a decision tree142def print_tree(node, depth=0):143    if isinstance(node, dict):144145        print('%s[X%s < %.3f]' % ((depth * ' ', (node['index'] + 1), float(node['value']))))146        print_tree(node['left'], depth + 1)147        print_tree(node['right'], depth + 1)148    else:149        print('%s[%s]' % ((depth * ' ', node)))150151def prun(copy_tree, depth, acc, dataset, class_val,copy_tree1,dir,acc_count):152153    if (copy_tree['right']==class_val[0] or  copy_tree['right']==class_val[1]) and( copy_tree['left']==class_val[0] or  copy_tree['left']==class_val[1]):154        # copy_tree['right']='None'155        # print('----------------------------------opt')156        # print_tree(copy_tree)157        if (int(copy_tree['count_left'])>int(copy_tree['count_right'])) or int(copy_tree['count_left'])==int(copy_tree['count_right']):158            if dir==2 or dir==0:159                copy_tree1['right'] = copy_tree1['right']['left']160            else:161                copy_tree1['left'] = copy_tree1['left']['left']162163        else:164            if dir == 2 or dir==0:165                copy_tree1['right'] = copy_tree1['right']['right']166            else:167                copy_tree1['left'] = copy_tree1['left']['right']168    elif (copy_tree['right']==class_val[0] or  copy_tree['right']==class_val[1] ) and isinstance(copy_tree, dict):169        # print("new left------------------------------------------------------------------")170        # print_tree(copy_tree)171        copy_tree1 = copy_tree172        prun(copy_tree['left'], depth - 1, acc, dataset, class_val, copy_tree1,1,acc_count)173    elif (copy_tree['left']==class_val[0] or  copy_tree['left']==class_val[1] ) and isinstance(copy_tree, dict):174        # print("new- right-----------------------------------------------------------------")175        # print_tree(copy_tree)176        copy_tree1 = copy_tree177        prun(copy_tree['right'], depth - 1, acc, dataset, class_val, copy_tree1,2,acc_count)178    elif isinstance(copy_tree, dict):179        # print("new------------------------------------------------------------------")180        # print_tree(copy_tree)181        copy_tree1=copy_tree182        prun(copy_tree['right'],depth-1,acc,dataset,class_val,copy_tree1,0,acc_count)183184    accura_prun=accuracy(dataset,copy_tree)185    while accura_prun >= acc:186        if accura_prun==acc:187            acc_count+=1188189        # print("new ittr")190        # print("new acc", accura_prun)191        if acc_count==2:192            break193        else:194            prun(copy_tree,depth_c,acc,dataset,class_val,copy_tree1,0,acc_count)195    return copy_tree1,accura_prun196197def accuracy(dataset,tree):198    correct_preditct = 0199    wrong_preditct = 0200    accuracy = 0201    for row in dataset:202        prediction = predict(tree, row)203        if prediction == row[-1]:204            correct_preditct += 1205        else:206            wrong_preditct += 1207        # print('Expected=%s, Got=%s' % (row[-1], prediction))208    accuracy = correct_preditct / (correct_preditct + wrong_preditct)209    # print("accuracy", accuracy * 100)210    return accuracy*100211212filename = 'dt_data.csv'213dataset = list()214with open(filename, 'r') as file:215    csv_reader = reader(file)216    for row in csv_reader:217        if not row:218            continue219        dataset.append(row)220dataset.pop(0)221222class_val= list(set(row[-1] for row in dataset))223tree = build_tree(dataset, depth_c,1 )224print_tree(tree)225accur=accuracy(dataset,tree)226print("accuracy of a full tree",accur)227Copy_tree=tree228# print(tree)229p_tree,acc_prun=prun(tree,depth_c,accur,dataset,class_val,Copy_tree,0,0)230231print("----------------------------------------------prun_tree")232print_tree(p_tree)233print("accuracy of pruned tree",acc_prun)234235# print(p_tree)236X=pd.DataFrame(dataset).iloc[:,:-1]237y=pd.DataFrame(dataset).iloc[:,-1]238239240# print("class_val",class_val)241for row in dataset:242    if (class_val[0]==row[2]):243        plt.scatter(row[0],row[1],color="red")244    else:245        plt.scatter(row[0],row[1],color="blue")246plt.show()247248print("best split")249print(gini_in)250251#252# min1, max1 = float(X.iloc[:, 0].min())-1, float(X.iloc[:, 0].max())+1253# min2, max2 = float(X.iloc[:, 1].min())-1, float(X.iloc[:, 1].max())+1254# print("min1",min1)255# print("max1",max1)256# print("min2",min2)257# print("max2",max2)258# x1grid = arange(min1, max1, 1)259# print(x1grid.shape)260# x2grid = arange(min2, max2, 0.5)261# print(x2grid.shape)262# xx, yy = meshgrid(x1grid, x2grid)263# r1, r2 = xx.flatten(), yy.flatten()264# r1, r2 = r1.reshape((len(r1), 1)), r2.reshape((len(r2), 1))265# grid = hstack((r1,r2))266# pred=[]267# for row in dataset:268#     pred.append(predict(tree,row))269# zz = np.array(pred).reshape(-1,1)270# # reshape(xx.shape)271# plt.contourf(xx, yy, zz, cmap='Paired')272# for class_value in range(1):273# 	row_ix = where(y == class_value)274# 	# create scatter of these samples275# 	plt.scatter(X[row_ix, 0], X[row_ix, 1], cmap='Paired')
...

Iris.py

Source:Iris.py

1from _csv import reader2import matplotlib.pyplot as plt3import pandas as pd456gini_in={}7info_gain={}8b_scoreG=[]9b_valueG=[]10depth_c=311count_acc=012def split_lr(index, value, dataset):13    left, right = list(), list()14    for row in dataset:15        if row[index] < value:16            left.append(row)17        else:18            right.append(row)19    return left, right202122# Gini_index calculating23def gini_impurity(groups, classes):24    n_instances = float(sum([len(group) for group in groups]))25    gini = 0.026    for group in groups:27        size = float(len(group))28        if size == 0:29            continue30        score = 0.031        for class_val in classes: # calculating the GINI Impurity32            p = [row[-1] for row in group].count(class_val) / size33            score += p * p34        gini += (1.0 - score) * (size / n_instances)35    if len(gini_in) != 0: #calculating the information gain36        gini = gini_in[b_valueG[len(b_valueG)-1]]-gini3738    return gini,score394041# Spliting the dataset42def spliting(dataset):43    class_count_left=044    class_count_right=045    class_values = list(set(row[-1] for row in dataset))46    row_index, row_value, gini_score, split_groups,score_t = 999, 999, 999, None,None47    for index in range(len(dataset[0]) - 1):48        for row in dataset:49            groups =split_lr(index, row[index], dataset)50            gini , score= gini_impurity(groups, class_values)51            if len(gini_in)==0:52                if gini < gini_score: # for the root node53                    row_index, row_value, gini_score, split_groups,score_t  = index, row[index], gini, groups,score54            else:55                gini_score=0 #getting the values of information gain for each split and selecting the once with the higher value56                if gini>gini_score:57                    row_index, row_value, gini_score, split_groups, score_t = index, row[index], gini, groups, score5859    gini_in[row_value]=gini_score60    b_scoreG.append(score_t)61    b_valueG.append(row_value)62    # if len(gini_in)==0:63    #     info_gain[]64    # print("gini",gini_in[b_scoreG[0]])65    left,right=split_groups6667    # outcomes = [row[-1] for row in left]68    # class_count_left=outcomes.count()69    # outcomes = [row[-1] for row in right]70    # class_count_right = outcomes.count()7172    return {'index': row_index, 'value': row_value, 'groups': split_groups,'count_left':0,'count_right':0}737475#terminal node76def to_terminal(group):77    outcomes = [row[-1] for row in group]78    count=079    for i in outcomes:80        if i==max(set(outcomes), key=outcomes.count):81            count+=182    return max(set(outcomes), key=outcomes.count),count838485# Create child splits for a node or make terminal, pruning based on the max depth and minimum size is done in this function86def split(node, max_depth, min_size, depth):87    left, right = node['groups']88    del (node['groups'])89    # check for a no split90    if not left or not right:91        val_left,counter=to_terminal(left + right)92        node['left'],node['count_left'] = node['right'],node['count_right'] = val_left,counter93        return94    # check for max depth95    if depth >= max_depth:96        val_left, counter_left = to_terminal(right)97        val_right,counter_right=to_terminal(right)98        node['left'],node['count_left'], node['right'],node['count_right'] = val_left, counter_left,val_right,counter_right99        return100    # process left child101    if len(left) <= min_size:102        val_left, counter_left = to_terminal(right)103        node['left'],node['count_left'] = val_left, counter_left104    else:105        node['left'] = spliting(left)106        split(node['left'], max_depth, min_size, depth + 1)107    # process right child108    if len(right) <= min_size:109        val_right, counter_right = to_terminal(right)110        node['right'],node['counter_right'] =  val_right, counter_right111    else:112        node['right'] = spliting(right)113        split(node['right'], max_depth, min_size, depth + 1)114115116# Build a decision tree117def build_tree(train, max_depth, min_size):118    root = spliting(train)119    split(root, max_depth, min_size, 1)120    return root121122def predict(node, row):123    test = float(row[node['index']])124    if  float(row[node['index']]) < float(node['value']):125        if isinstance(node['left'], dict):126            return predict(node['left'], row)127        else:128            return node['left']129    else:130        if isinstance(node['right'], dict):131            return predict(node['right'], row)132        else:133            return node['right']134135136# Print a decision tree137def print_tree(node, depth=0):138    if isinstance(node, dict):139140        print('%s[X%s < %.3f]' % ((depth * ' ', (node['index'] + 1), float(node['value']))))141        print_tree(node['left'], depth + 1)142        print_tree(node['right'], depth + 1)143    else:144        print('%s[%s]' % ((depth * ' ', node)))145146def prun(copy_tree, depth, acc, dataset, class_val,copy_tree1,dir,acc_count):147148    if (copy_tree['right']==class_val[0] or  copy_tree['right']==class_val[1] or copy_tree['right']==class_val[2]) and( copy_tree['left']==class_val[0] or  copy_tree['left']==class_val[1] or copy_tree['left']==class_val[2]):149        # copy_tree['right']='None'150        # print('----------------------------------opt')151        # print_tree(copy_tree)152        if (int(copy_tree['count_left'])>int(copy_tree['count_right'])) or int(copy_tree['count_left'])==int(copy_tree['count_right']):153            if dir==2 or dir==0:154                copy_tree1['right'] = copy_tree1['right']['left']155            else:156                copy_tree1['left'] = copy_tree1['left']['left']157158        else:159            if dir == 2 or dir==0:160                copy_tree1['right'] = copy_tree1['right']['right']161            else:162                copy_tree1['left'] = copy_tree1['left']['right']163    elif (copy_tree['right']==class_val[0] or  copy_tree['right']==class_val[1] or copy_tree['right']==class_val[2]) and isinstance(copy_tree, dict):164        # print("new left------------------------------------------------------------------")165        # print_tree(copy_tree)166        copy_tree1 = copy_tree167        prun(copy_tree['left'], depth - 1, acc, dataset, class_val, copy_tree1,1,acc_count)168    elif (copy_tree['left']==class_val[0] or  copy_tree['left']==class_val[1] or copy_tree['left']==class_val[2]) and isinstance(copy_tree, dict):169        # print("new- right-----------------------------------------------------------------")170        # print_tree(copy_tree)171        copy_tree1 = copy_tree172        prun(copy_tree['right'], depth - 1, acc, dataset, class_val, copy_tree1,2,acc_count)173    elif isinstance(copy_tree, dict):174        # print("new------------------------------------------------------------------")175        # print_tree(copy_tree)176        copy_tree1=copy_tree177        prun(copy_tree['right'],depth-1,acc,dataset,class_val,copy_tree1,0,acc_count)178179    accura_prun=accuracy(dataset,copy_tree)180    while accura_prun >= acc:181        if accura_prun==acc:182            acc_count+=1183184        # print("new ittr")185        # print("new acc", accura_prun)186        if acc_count==2:187            break188        else:189            prun(copy_tree,depth_c,acc,dataset,class_val,copy_tree1,0,acc_count)190    return copy_tree1,accura_prun191192def accuracy(dataset,tree):193    correct_preditct = 0194    wrong_preditct = 0195    accuracy = 0196    for row in dataset:197        prediction = predict(tree, row)198        if prediction == row[-1]:199            correct_preditct += 1200        else:201            wrong_preditct += 1202        # print('Expected=%s, Got=%s' % (row[-1], prediction))203    accuracy = correct_preditct / (correct_preditct + wrong_preditct)204    # print("accuracy", accuracy * 100)205    return accuracy*100206207filename = 'Iris.csv'208dataset = list()209with open(filename, 'r') as file:210    csv_reader = reader(file)211    for row in csv_reader:212        if not row:213            continue214        dataset.append(row)215dataset.pop(0)216217class_val= list(set(row[-1] for row in dataset))218tree = build_tree(dataset, depth_c,1 )219print_tree(tree)220accur=accuracy(dataset,tree)221print("accuracy of a full tree",accur)222Copy_tree=tree223# print(tree)224p_tree,acc_prun=prun(tree,depth_c,accur,dataset,class_val,Copy_tree,0,0)225226print("----------------------------------------------prun_tree")227print_tree(p_tree)228print("accuracy of pruned tree",acc_prun)229230# print(p_tree)231X=pd.DataFrame(dataset).iloc[:,:-1]232y=pd.DataFrame(dataset).iloc[:,-1]233234# print("class_val",class_val)235plt.figure(figsize=(12,12))236for row in dataset:237    if (class_val[0]==row[-1]):238        plt.scatter(row[0],row[1],color="red")239    elif (class_val[1]==row[-1]):240        plt.scatter(row[0],row[1],color="blue")241    else:242        plt.scatter(row[0], row[1], color="black")243plt.show()244245print("best_split")
...

CreateEffekseerFiles_v4.py

Source:CreateEffekseerFiles_v4.py

1import shutil2import os3from distutils.dir_util import copy_tree4copy_tree("../Effekseer/Dev/Cpp/Effekseer", "./Players/Cocos2d-x_v4/Effekseer")5copy_tree("../Effekseer/Dev/Cpp/EffekseerMaterialCompiler", "./Players/Cocos2d-x_v4/EffekseerMaterialCompiler")6copy_tree("../Effekseer/Dev/Cpp/EffekseerRendererCommon", "./Players/Cocos2d-x_v4/EffekseerRendererCommon")7copy_tree("../Effekseer/Dev/Cpp/EffekseerRendererGL", "./Players/Cocos2d-x_v4/EffekseerRendererGL")8copy_tree("../Effekseer/Dev/Cpp/EffekseerRendererLLGI", "./Players/Cocos2d-x_v4/EffekseerRendererLLGI")9copy_tree("../Effekseer/Dev/Cpp/EffekseerRendererMetal", "./Players/Cocos2d-x_v4/EffekseerRendererMetal")10os.makedirs("./Players/Cocos2d-x_v4/LLGI/", exist_ok=True)11shutil.copyfile("../Effekseer/Dev/Cpp/3rdParty/LLGI/CMakeLists.txt", "./Players/Cocos2d-x_v4/3rdParty/LLGI/CMakeLists.txt")12copy_tree("../Effekseer/Dev/Cpp/3rdParty/LLGI/src", "./Players/Cocos2d-x_v4/3rdParty/LLGI/src")13copy_tree("../Effekseer/Dev/Cpp/3rdParty/LLGI/cmake", "./Players/Cocos2d-x_v4/3rdParty/LLGI/cmake")...

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.