12.14

Author: Naplues

main.py

@@ -41,8 +41,10 @@ def run_default():
     # ======= NLP-based approaches ======= NGram, NGram_C
     # ======= Glance-XX approaches ======= Glance_MD, Glance_EA, Glance_LR
 
-    run_cross_release_predict(Glance_LR)
-
+    # run_cross_release_predict(Glance_LR)
+    # run_cross_release_predict(Glance_EA)
+    # run_cross_release_predict(Glance_MD)
+    run_cross_release_predict(LineDP)
 
 def parse_args():
     # If there is no additional parameters in the command line, run the default models.

result/Dis1_1/effort-total.csv

@@ -0,0 +1,124 @@
+release,total_effort
+ambari-2.1.0,292084
+ambari-2.2.0,316411
+ambari-2.4.0,445436
+ambari-2.5.0,484406
+ambari-2.6.0,487664
+ambari-2.7.0,484089
+amq-5.1.0,190852
+amq-5.2.0,204975
+amq-5.4.0,260959
+amq-5.5.0,274603
+amq-5.6.0,306066
+amq-5.7.0,305607
+amq-5.8.0,311711
+amq-5.9.0,315307
+amq-5.10.0,323082
+amq-5.11.0,333415
+amq-5.12.0,350040
+amq-5.14.0,370472
+amq-5.15.0,373372
+bookkeeper-4.2.0,78715
+bookkeeper-4.4.0,60408
+calcite-1.8.0,327252
+calcite-1.11.0,341392
+calcite-1.13.0,273186
+calcite-1.15.0,288305
+calcite-1.16.0,295153
+calcite-1.17.0,299074
+calcite-1.18.0,308933
+cassandra-0.8.6,131184
+cassandra-1.0.9,144502
+cassandra-1.1.6,168159
+cassandra-1.1.11,170533
+cassandra-1.2.11,211949
+flink-1.6.0,608219
+groovy-1.5.5,109318
+groovy-1.6.0,141955
+groovy-1.7.3,156480
+groovy-1.7.6,159176
+groovy-1.8.1,176662
+groovy-1.8.7,183462
+groovy-2.1.0,203909
+groovy-2.1.6,206046
+groovy-2.4.4,243567
+groovy-2.4.6,244632
+groovy-2.4.8,245541
+groovy-2.5.0,279517
+groovy-2.5.5,280363
+hbase-0.94.5,280440
+hbase-0.98.0,606913
+hbase-0.98.5,623281
+hbase-0.98.11,654904
+hive-1.2.0,871574
+hive-2.0.0,1039823
+hive-2.1.0,1112941
+ignite-1.4.0,610699
+ignite-1.6.0,690475
+log4j2-2.1,79796
+log4j2-2.2,80161
+log4j2-2.3,80883
+log4j2-2.4,93481
+log4j2-2.5,103402
+log4j2-2.6,121815
+log4j2-2.7,132836
+log4j2-2.8,138778
+log4j2-2.9,145534
+log4j2-2.10,148721
+mahout-0.4,125301
+mahout-0.5,124968
+mahout-0.6,145593
+mahout-0.7,133660
+mahout-0.8,149368
+mng-3.1.0,91695
+mng-3.2.0,91886
+mng-3.3.0,98343
+mng-3.5.0,99215
+mng-3.6.0,100401
+nifi-1.2.0,416418
+nifi-1.5.0,489997
+nifi-1.8.0,560333
+nutch-1.3,45486
+nutch-1.4,44989
+nutch-1.5,46006
+nutch-1.6,46843
+nutch-1.7,47477
+nutch-1.8,48420
+nutch-1.9,47780
+nutch-1.10,55704
+nutch-1.12,69185
+nutch-1.13,70459
+nutch-1.14,72871
+nutch-1.15,77351
+storm-0.9.3,80758
+storm-1.0.0,221337
+storm-1.0.3,234518
+storm-1.0.5,235508
+tika-0.8,30890
+tika-0.9,34174
+tika-0.10,46748
+tika-1.1,50768
+tika-1.3,58365
+tika-1.5,61281
+tika-1.7,69500
+tika-1.10,87547
+tika-1.13,93972
+tika-1.15,117323
+tika-1.17,120209
+ww-2.0.5,69404
+ww-2.0.10,74189
+ww-2.1.1,75656
+ww-2.1.3,84193
+ww-2.1.7,88316
+ww-2.2.0,195385
+ww-2.2.2,197224
+ww-2.3.1,198580
+ww-2.3.4,199575
+ww-2.3.10,200322
+ww-2.3.15,202082
+ww-2.3.17,204101
+ww-2.3.20,205872
+ww-2.3.24,207064
+zookeeper-3.5.1,65270
+zookeeper-3.5.2,66505
+zookeeper-3.5.3,68039

result/Dis1/Dis1-summary-mean-Glance-EA-horizontal.csv result/Dis1_2/Dis1-summary-mean-Glance-EA-horizontal.csv

@@ -0,0 +1,55 @@
+# -*- coding:utf-8 -*-
+
+import sys
+
+import warnings
+import pandas as pd
+from pandas import DataFrame
+
+sys.path.append('C:/Users/gzq-712/Desktop/Git/CLDP/')
+from src.models.glance import *
+from statistics import *
+
+output_path = '../../result/Dis1_1/'
+
+
+def get_effort_of_Glance(clf):
+    print(clf)
+    total_data = read_data_from_file(f'{output_path}effort-total.csv')[1:]
+    text = 'release,review_effort,total_effort,ratio\n'
+    count = 0
+    for project, releases in get_project_releases_dict().items():
+        for i in range(len(releases) - 1):
+            total_effort = int(total_data[count].split(',')[1])
+            file_path = f'{root_path}Result/{clf}/line_result/{project}/{releases[i + 1]}-result.csv'
+            review_effort = len(read_data_from_file(file_path)) - 1
+            text += f'{releases[i + 1]},{review_effort},{total_effort},{round(review_effort / total_effort, 3)}\n'
+            count += 1
+    save_csv_result(output_path, f'effort-{clf}.csv', text)
+
+
+def summary():
+    MD = read_data_from_file(f'{output_path}effort-Glance-MD.csv')[1:]
+    EA = read_data_from_file(f'{output_path}effort-Glance-EA.csv')[1:]
+    LR = read_data_from_file(f'{output_path}effort-Glance-LR.csv')[1:]
+
+    text = 'release,MD_effort,EA_effort,LR_effort\n'
+    count = 0
+    for project, releases in get_project_releases_dict().items():
+        MD_data, EA_data, LR_data = [], [], []
+        for i in range(len(releases) - 1):
+            MD_data.append(float(MD[count].split(',')[3]))
+            EA_data.append(float(EA[count].split(',')[3]))
+            LR_data.append(float(LR[count].split(',')[3]))
+            count += 1
+        text += f'{project},{median(MD_data)},{median(EA_data)},{median(LR_data)}\n'
+
+    save_csv_result(output_path, f'effort-comparison.csv', text)
+
+
+if __name__ == '__main__':
+    for clf in ['Glance-MD', 'Glance-EA', 'Glance-LR']:
+        # get_effort_of_Glance(clf)
+        pass
+    summary()
+    pass

result/Dis1/Dis1-summary-mean-Glance-EA-vertical.csv result/Dis1_2/Dis1-summary-mean-Glance-EA-vertical.csv

@@ -19,7 +19,7 @@
                      '75%', '80%', '85%', '90%', '95%', '100%', ]
 indicators = ['recall', 'far', 'ce', 'd2h', 'mcc', 'ifa', 'recall_20', 'ratio']
 
-output_path = '../../result/Dis1/'
+output_path = '../../result/Dis1_2/'
 make_path(output_path)
 
 

result/Dis1/Dis1-summary-mean-Glance-LR-horizontal.csv result/Dis1_2/Dis1-summary-mean-Glance-LR-horizontal.csv

@@ -0,0 +1,65 @@
+# -*- coding: utf-8 -*-
+
+import sys
+
+sys.path.append('C:/Users/gzq-712/Desktop/Git/CLDP/')
+from statistics import mean, median
+import pandas as pd
+from pandas import DataFrame
+
+from src.models.glance import *
+
+output_path = '../../result/Dis1_3/'
+
+
+def select_models():
+    """
+    Select proper models according to different purposes.
+    :param exp: Experiment name
+    :return: A list of model instances.
+    """
+    # , Glance_B_NT_noCC(), Glance_B_NFC_noCC
+    return [Glance_LR(), Glance_B_noCC(), Glance_B_NT(), Glance_B_NT_noCC(), Glance_B_NFC(), Glance_B_NFC_noCC()]
+
+
+def collect_line_level_summary_result(eva_method=None):
+    if eva_method is None:
+        eva_method = [mean, median]
+    text = ''
+    for method in eva_method:
+        text += f'Approach,Recall,FAR,CE,D2H,MCC,IFA,Recall@20%,ratio\n'
+        for model in select_models():
+            df = pd.read_csv(model.line_level_evaluation_file)
+
+            recall = round(method(list(df['recall'])), 3)
+            far = round(method(list(df['far'])), 3)
+            ce = round(method(list(df['ce'])), 3)
+            d2h = round(method(list(df['d2h'])), 3)
+            mcc = round(method(list(df['mcc'])), 3)
+            ifa = int(method(list(df['ifa'])))
+            recall_20 = round(method(list(df['recall_20'])), 3)
+            ratio = round(method(list(df['ratio'])), 3)
+            # ER = round(method(list(df['ER'])), 3)
+            # RI = round(method(list(df['RI'])), 3)
+            text += f'{model.model_name},{recall},{far},{ce},{d2h},{mcc},{ifa},{recall_20},{ratio}\n'
+        text += '\n'
+    save_csv_result(output_path, f'Performance_Summary.csv', text)
+
+
+# =================== Line level result in terms of different Performance Indicators experiments ================
+def collect_line_level_by_indicators():
+    models = select_models()
+    indicators = ['recall', 'far', 'ce', 'd2h', 'mcc', 'ifa', 'recall_20', 'ratio']
+    for indicator in indicators:
+        data = dict()
+        for model in models:
+            data[model.model_name] = pd.read_csv(model.line_level_evaluation_file)[indicator].tolist()
+
+        ratio = DataFrame(data, columns=[model.model_name for model in models])
+        ratio.to_csv(f'{output_path}Performance Indicators/{indicator}.csv', index=False)
+
+
+if __name__ == '__main__':
+    collect_line_level_summary_result()
+    collect_line_level_by_indicators()
+    pass

result/Dis1/Dis1-summary-mean-Glance-LR-vertical.csv result/Dis1_2/Dis1-summary-mean-Glance-LR-vertical.csv

@@ -81,44 +81,6 @@ def classification_difference(exp="RQ2"):
     save_csv_result(f'../../result/{exp}/', 'Difference_summary.csv', text)
 
 
-def diff_ranking():
-    access_model = 'AccessModel'
-    target_model = 'CheckStyleModel'  # PMDModel CheckStyleModel
-    text = ''
-    for project, releases in get_project_releases_dict().items():
-
-        print(project, end=',')
-        increase, decrease = [], []
-        for release in releases[1:]:
-            data_access = load_pk_result(f'{result_path}Diff_Ranking/{access_model}/{release}.pk')
-            data_target = load_pk_result(f'{result_path}Diff_Ranking/{target_model}/{release}.pk')
-            i, d = [], []
-            t, text_lines, label, filename = read_file_level_dataset(release)
-            len_dict = {}
-            for index in range(len(filename)):
-                len_dict[filename[index]] = len(text_lines[index])
-
-            for file_line in data_target.keys():
-                if file_line not in data_access.keys():
-                    continue
-                rank_of_access = data_access[file_line]
-                rank_of_target = data_target[file_line]
-                # increase
-                if rank_of_access < rank_of_target:
-                    i.append((rank_of_target - rank_of_access) / len_dict[file_line.split(':')[0]])
-                # decrease
-                if rank_of_access > rank_of_target:
-                    d.append((rank_of_access - rank_of_target) / len_dict[file_line.split(':')[0]])
-
-            increase.append(mean(i))
-            decrease.append(mean(d))
-
-        text += project + ',' + str(increase).replace('[', '').replace(']', '') + '\n'
-        text += project + ',' + str(decrease).replace('[', '').replace(']', '') + '\n'
-    print(text)
-    save_result('C:/Users/GZQ/Desktop/d.csv', text)
-
-
 if __name__ == '__main__':
     # diff_ranking()
     for experiment in ["RQ2", "RQ3"]:

result/Dis1/Dis1-summary-mean-Glance-MD-horizontal.csv result/Dis1_2/Dis1-summary-mean-Glance-MD-horizontal.csv

@@ -242,15 +242,16 @@ def analyze_line_level_result(self):
             ifa, r_20 = self.rank_strategy_3()  # Strategy 1
 
         ################################ Bug hit ratio ################################################
-        buggy_lines_dict = read_dict_from_file(f'{self.commit_buggy_path}/{self.test_release}_commit_buggy_lines.csv')
+        # buggy_lines_dict = read_dict_from_file(f'{self.commit_buggy_path}/{self.test_release}_commit_buggy_lines.csv')
+        buggy_lines_dict = {}  # = read_dict_from_file(f'{self.commit_buggy_path}/{self.test_release}_commit_buggy_lines.csv')
         total_bugs = len(buggy_lines_dict.keys())
         hit_bugs = set()
         for line in self.predicted_buggy_lines:
             for bug_commit, lines in buggy_lines_dict.items():
                 if line in lines:
                     hit_bugs.add(bug_commit)
 
-        ratio = round(len(hit_bugs) / total_bugs, 3)
+        ratio = 0 if total_bugs == 0 else round(len(hit_bugs) / total_bugs, 3)
 
         ################################ Output the evaluation result ################################################
         append_title = True if not os.path.exists(self.line_level_evaluation_file) else False