Merge pull request Codium-ai#18 from Cassini-chris/patch-2

Optimized and commented code

alpha_codium/evaluate_dataset.py

@@ -7,25 +7,31 @@
 
 logger = get_logger(__name__)
 
-
 def evaluate_dataset_solution(dataset_name='valid_and_test_processed',
-                              split_name='test',
-                              solution_path_database='valid_database_solution.json'):
-    split_name = split_name
-    dataset_name = dataset_name
-    data_provider = CodeContestDataProvider(dataset_location=dataset_name)
-
-    ds = data_provider.dataset[split_name]
-
-    solution_path_database = solution_path_database
+                               split_name='test',
+                               solution_path_database='valid_database_solution.json'):
+    """
+    Evaluate the performance of dataset solutions.
 
+    Args:
+        dataset_name (str, optional): The name of the dataset. Defaults to 'valid_and_test_processed'.
+        split_name (str, optional): The name of the split. Defaults to 'test'.
+        solution_path_database (str, optional): The path to the solution database file. Defaults to 'valid_database_solution.json'.
+    """
 
+    # Load the dataset and solution database
+    data_provider = CodeContestDataProvider(dataset_location=dataset_name)
+    ds = data_provider.dataset[split_name]
     with open(solution_path_database, 'r') as f:
         database_solutions = json.load(f)
         database_solutions[split_name] = OrderedDict(
             sorted(database_solutions[split_name].items(), key=lambda x: int(x[0])))
+
+    # Initialize counters for passed and failed problems
     total_passed = 0
     total_failed = 0
+
+    # Iterate over the solutions in the database
     for sol in database_solutions[split_name]:
         try:
             key_str = sol
@@ -68,9 +74,12 @@ def evaluate_dataset_solution(dataset_name='valid_and_test_processed',
             print(f"Error: {e}")
             pass
 
+    # Print the total number of passed and failed problems
     print(f"total_passed: {total_passed}, total_failed: {total_failed}")
-    print(f"pass rate: {total_passed/(total_passed+total_failed)}")
 
+    # Calculate the pass rate
+    pass_rate = total_passed / (total_passed + total_failed)
+    print(f"pass rate: {pass_rate}")
 
 parser = argparse.ArgumentParser()
 parser.add_argument("--dataset_name", type=str, default="valid_and_test_processed")