Update src/populate.py

src/populate.py

@@ -12,23 +12,31 @@ def get_leaderboard_df(results_path: str, requests_path: str, cols: list, benchm
     """Creates a dataframe from all the individual experiment results"""
     raw_data = get_raw_eval_results(results_path, requests_path)
     all_data_json = [v.to_dict() for v in raw_data]
+    
+    print(f"All data JSON: {all_data_json}")
+
 
 
     df = pd.DataFrame.from_records(all_data_json)
+    print(f"Initial DataFrame: {df}")
     score_cols = [
         'ALT E to J BLEU', 'ALT J to E BLEU', 'WikiCorpus E to J BLEU', 'WikiCorpus J to E BLEU', 
         'XL-Sum JA BLEU', 'XL-Sum ROUGE1', 'XL-Sum ROUGE2', 'XL-Sum ROUGE-Lsum'
     ]
     
     existing_score_cols = [col for col in score_cols if col in df.columns]
+    print(f"Existing score columns: {existing_score_cols}")
     
     # スコア列を100で割り、.4f形式でフォーマット
     df[existing_score_cols] = (df[existing_score_cols] / 100).applymap(lambda x: f'{x:.4f}')
+    print(f"DataFrame after score adjustment: {df}")
     df = df.sort_values(by=[AutoEvalColumn.AVG.name], ascending=False)
     df = df[cols].round(decimals=2)
-
+    print(f"Sorted DataFrame: {df}")
+    
     # filter out if any of the benchmarks have not been produced
     df = df[has_no_nan_values(df, benchmark_cols)]
+    print(f"Final DataFrame after NaN filtering: {df}")
     return df