Update

app.py

@@ -54,10 +54,11 @@ for bug_id, time in bug_id_to_time.items():
     timeline_xs.append(time)
     timeline_ys.append(0)
     timeline_cols.append("Baseline")
-LEADERBOARD_DF = get_leaderboard_df(EVAL_REQUESTS_PATH, COLS, total_issues)
+LEADERBOARD_DF = get_leaderboard_df(EVAL_REQUESTS_PATH, total_issues)
 for row in LEADERBOARD_DF.itertuples():
+    print(row)
     model_cnt += 1
-    for fix in json.loads(row.fixed_bug_ids):
+    for fix in row.fixed_bug_ids:
         timeline_xs.append(bug_id_to_time[fix])
         timeline_ys.append(model_cnt)
         timeline_cols.append(row.method_name)

src/leaderboard/read_evals.py

@@ -94,7 +94,7 @@ class EvalResult:
                 (self.build_count - self.build_failure_count) * 100.0 / self.build_count, 1
             ),
             AutoEvalColumn.mttr.name: self.mttr,
-            "fixed_bug_ids": json.dumps(self.fixed_bug_ids),
+            "fixed_bug_ids": self.fixed_bug_ids,
             "method_id": self.method_name + "(" + self.model_name + ")",
         }
 

src/populate.py

@@ -7,12 +7,11 @@ from src.display.utils import AutoEvalColumn
 from src.leaderboard.read_evals import get_raw_eval_results
 
 
-def get_leaderboard_df(requests_path: str, cols: list, total_issues: int) -> pd.DataFrame:
+def get_leaderboard_df(requests_path: str, total_issues: int) -> pd.DataFrame:
     """Creates a dataframe from all the individual experiment results"""
     raw_data = get_raw_eval_results(requests_path)
     all_data_json = [v.to_dict(total_issues) for v in raw_data]
 
     df = pd.DataFrame.from_records(all_data_json)
     df = df.sort_values(by=[AutoEvalColumn.full_pass_count.name], ascending=False)
-    df = df[cols].round(decimals=2)
     return df