add heat map

analysis/bpf/heatmap.py

@@ -0,0 +1,48 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+# Load the CSV file
+file_path = 'data/commit_survey.csv'   # Replace with the actual path to your CSV file
+df = pd.read_csv(file_path)
+
+# Keywords to filter commits
+keywords = ['bpf_link', 'dynptr', 'token', 'bpf_iter', 'tail_call', 'bpf_timer', 'spin_lock']
+
+# Filter out rows where the implementation component contains test cases or is marked as unrelated/not sure
+filtered_df = df[~df['major_related_implementation_component'].str.contains('test', case=False, na=False) &
+                 ~df['major_related_implementation_component'].str.contains('not related', case=False, na=False) &
+                 ~df['major_related_implementation_component'].str.contains('not sure', case=False, na=False)]
+
+# Create an empty DataFrame to store the counts of each keyword for each implementation component
+keyword_component_counts = pd.DataFrame()
+
+# Iterate through each keyword to calculate the counts of related implementation components
+for keyword in keywords:
+    keyword_commits = filtered_df[filtered_df['commit_message'].str.contains(keyword, case=False, na=False) | 
+                                  filtered_df['summary'].str.contains(keyword, case=False, na=False)]
+
+    keyword_implementation_component = keyword_commits[['major_related_implementation_component', 'commit_id']]
+    component_count = keyword_implementation_component.groupby('major_related_implementation_component').count()
+    component_count.columns = [keyword]
+
+    # Merge with the main DataFrame to accumulate the counts
+    if keyword_component_counts.empty:
+        keyword_component_counts = component_count
+    else:
+        keyword_component_counts = keyword_component_counts.merge(component_count, left_index=True, right_index=True, how='outer')
+
+# Fill any missing values with 0
+keyword_component_counts.fillna(0, inplace=True)
+
+# Limit the label length to a maximum of 20 characters for readability
+keyword_component_counts.index = keyword_component_counts.index.str[:20]
+
+# Create the heatmap
+plt.figure(figsize=(8, 5))
+sns.heatmap(keyword_component_counts, annot=True, fmt="g", cmap='Blues', cbar=True)
+# plt.title("Heatmap of BPF Link, Dynptr, Token, BPF Iter, Tail Call, BPF Timer, and Spin Lock Commits vs. Implementation Components")
+plt.ylabel("Implementation Component")
+plt.xlabel("Keywords")
+plt.tight_layout()
+plt.savefig('imgs/heatmap_bpf_keywords_vs_components.png')