Update

app.py

@@ -50,10 +50,15 @@ bug_id_by_cat = {
     "miscompilation": [],
     "hang": [],
 }
+bug_id_to_comp = dict()
+comp_bug_count = dict()
 for issue in dataset["test"]:
     bug_id_to_time[issue["bug_id"]] = pd.to_datetime(issue["knowledge_cutoff"])
     bug_id_by_cat[issue["bug_type"]].append(issue["bug_id"])
     bug_id_to_type[issue["bug_id"]] = issue["bug_type"]
+    bug_id_to_comp[issue["bug_id"]] = issue["components"]
+    for comp in issue["components"]:
+        comp_bug_count[comp] = comp_bug_count.get(comp, 0) + 1
 timeline_xs = []
 timeline_ys = []
 timeline_cols = []
@@ -107,6 +112,21 @@ fixed_by_cat_df = pd.DataFrame(
     }
 )
 fixed_by_cat_df.sort_values("Total", inplace=True, ascending=False)
+fixed_by_comp = dict()
+for bug_id in fixed_bug_ids:
+    for comp in bug_id_to_comp[bug_id]:
+        fixed_by_comp[comp] = fixed_by_comp.get(comp, 0) + 1
+fixed_by_comp_df = pd.DataFrame(
+    {
+        "Component": [comp for comp in fixed_by_comp.keys()],
+        "Total": [comp_bug_count[comp] for comp in fixed_by_comp.keys()],
+        "Repaired": list(fixed_by_comp.values()),
+        "Repair Rate (%)": [
+            round(fixed_by_comp[comp] / comp_bug_count[comp] * 100, 1) for comp in fixed_by_comp.keys()
+        ],
+    }
+)
+fixed_by_comp_df.sort_values("Total", inplace=True, ascending=False)
 
 
 def init_leaderboard(dataframe):
@@ -150,6 +170,7 @@ with demo:
                 tooltip=["bug_id", "method_name", "time", "bug_type"],
             )
             gr.Dataframe(fixed_by_cat_df)
+            gr.Dataframe(fixed_by_comp_df)
 
         with gr.TabItem("🚀 Submission", elem_id="llm-benchmark-tab-table", id=1):
             gr.Markdown(EVALUATION_QUEUE_TEXT, elem_classes="markdown-text")