From 35f071c9c0c57259199dca931b5ba4b165d314b2 Mon Sep 17 00:00:00 2001
From: Thomas Schweizer <tschweiz@cs.washington.edu>
Date: Sun, 10 Dec 2023 21:52:28 -0800
Subject: [PATCH] Add ground truth size script

---
 .../main/analysis/print_ground_truth_size.py  | 65 +++++++++++++++++++
 src/python/main/metrics.py                    |  7 ++
 2 files changed, 72 insertions(+)
 create mode 100644 src/python/main/analysis/print_ground_truth_size.py
 create mode 100644 src/python/main/metrics.py

diff --git a/src/python/main/analysis/print_ground_truth_size.py b/src/python/main/analysis/print_ground_truth_size.py
new file mode 100644
index 0000000..5228ec5
--- /dev/null
+++ b/src/python/main/analysis/print_ground_truth_size.py
@@ -0,0 +1,65 @@
+"""
+Prints the ground truth sizes for each dataset as Latex commands in stdout.
+"""
+import sys
+import os
+
+sys.path.insert(1, os.path.join(sys.path[0], '..'))
+import metrics
+import evaluation_results
+import argparse
+
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+def main(d4j_ground_truth_file: str, lltc4j_ground_truth_file: str):
+    d4j_ground_truth = evaluation_results.read_ground_truth(d4j_ground_truth_file)
+    lltc4j_ground_truth = evaluation_results.read_ground_truth(lltc4j_ground_truth_file)
+
+    df1 = metrics.calculate_ground_truth_size_per_commit(d4j_ground_truth)
+    df1['dataset'] = 'Defects4J'
+
+    df2 = metrics.calculate_ground_truth_size_per_commit(lltc4j_ground_truth)
+    df2['dataset'] = 'LLTC4J'
+
+    result_df = pd.concat([df1, df2], ignore_index=True)[['count', 'dataset']]
+
+    # print(result_df)
+    # sns.histplot(result_df, x='count', hue='dataset', log_scale=True, element="step", fill=True)
+    # plt.savefig('ground_truth_size.png')
+
+    mean_df = result_df[['dataset', 'count']].groupby('dataset').agg(['mean', 'std']).round(2)
+
+
+    for index, row in mean_df.iterrows():
+        dataset_name_for_latex = index.lower().replace('4', 'f')
+        print(f"\\newcommand\\{dataset_name_for_latex}MeanSize{{{row['count']['mean']}\\xspace}}")
+        print(f"\\newcommand\\{dataset_name_for_latex}StdSize{{{row['count']['std']}\\xspace}}")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(
+        prog=sys.argv[0],
+        description=__doc__,
+        formatter_class=argparse.RawDescriptionHelpFormatter
+    )
+
+    parser.add_argument(
+        "--d4j-ground-truth",
+        help="Path to the ground truth file containing all Defects4J commits",
+        required=True,
+        metavar="D4J_GROUND_TRUTH_FILE",
+    )
+
+    parser.add_argument(
+        "--lltc4j-ground-truth",
+        help="Path to the ground truth file containing all LLTC4J commits",
+        required=True,
+        metavar="LLTC4J_GROUND_TRUTH_FILE",
+    )
+
+    args = parser.parse_args()
+    main(args.d4j_ground_truth, args.lltc4j_ground_truth)
+
+
diff --git a/src/python/main/metrics.py b/src/python/main/metrics.py
new file mode 100644
index 0000000..50b42aa
--- /dev/null
+++ b/src/python/main/metrics.py
@@ -0,0 +1,7 @@
+"""
+Metric utilities
+"""
+import pandas as pd
+
+def calculate_ground_truth_size_per_commit(dataframe : pd.DataFrame) -> pd.DataFrame:
+    return dataframe[['project', 'commit_id', 'file']].groupby(['project', 'commit_id']).count().reset_index().rename(columns= {'file':'count'})