Merge pull request #24 from Aignosi/fix/SIENTIAPDE-1579

Upgrade MLflow to 2.18.0 and constrain setuptools

model_manager/sientia/model_serving.py

@@ -56,7 +56,7 @@ class ModelServing:
     # Function to list runs for a given experiment
     def search_runs_by_name(
         self, experiment_names: list[str], order_by: None | list[str] = None
-    ) -> pd.DataFrame:
+    ) -> pd.DataFrame | list:
         """
         List runs for a specified MLflow experiment.
 
@@ -64,7 +64,7 @@ class ModelServing:
             experiment_names (list[str]): List with experiment_names to retrieve runs from.
 
         Returns:
-            pandas.DataFrame: A DataFrame containing run information.
+            Union[pd.DataFrame, list]: A DataFrame or list containing run information.
 
         Raises:
             SientiaMlException: If unable to search runs.
@@ -73,7 +73,7 @@ class ModelServing:
             runs = mlflow.search_runs(experiment_names=experiment_names, order_by=order_by)
         except SientiaMlException as e:
             logging.error(e)
-            raise SientiaMlException from e
+            raise SientiaMlException(str(e)) from e
         return runs
 
     def set_experiment(self, experiment_identifier: str) -> None:

model_manager/sientia/models.py

@@ -35,9 +35,7 @@ def validate_frontend_date_format(fmt: str | None) -> None:
         return
     if fmt not in ALLOWED_FRONTEND_DATE_FORMATS:
         allowed = ', '.join(sorted(ALLOWED_FRONTEND_DATE_FORMATS))
-        raise ValueError(
-            f'Invalid date_format "{fmt}". Allowed formats: {allowed}'
-        )
+        raise ValueError(f'Invalid date_format "{fmt}". Allowed formats: {allowed}')
 
 
 def _frontend_date_format_to_strftime(fmt: str | None) -> str | None:

requirements.txt

@@ -5,7 +5,8 @@ boto3==1.40.55
 botocore==1.40.55
 git+ssh://git@github.com/Aignosi/sientia-dataops-library.git@1.6.1
 prometheus-client==0.23.1
-mlflow==2.10.1
+mlflow==2.18.0
+setuptools<81
 evidently==0.4.21
 beautifulsoup4==4.12.3
 scikit-learn==1.4.2

scripts/run_training_test.py

@@ -51,6 +51,12 @@ TEST_SCENARIOS_DIR = PROJECT_ROOT / 'docs' / 'test-scenarios'
 MINIO_ALIAS = 'suse'
 MINIO_BUCKET = 'model-training'
 
+# Mapeamento de CSV específico por cenário
+SCENARIO_CSV_MAPPING = {
+    '12-angular-test-date-format': Path('docs/DB_CV022_WIT230.csv'),
+    '13-angular-test-double-date-column': Path('docs/DB_CV022_WIT230 _double_date_column.csv'),
+}
+
 POSTGRES_CONFIG = {
     'host': os.getenv('POSTGRES_HOST'),
     'port': os.getenv('POSTGRES_PORT'),
@@ -306,6 +312,10 @@ def run_local_pipeline(
         result['error'] = str(exc)
         return result
 
+    # Use scenario-specific CSV if mapped, otherwise use provided csv_path
+    if scenario_name in SCENARIO_CSV_MAPPING:
+        csv_path = SCENARIO_CSV_MAPPING[scenario_name]
+    
     csv_path = _resolve_csv_path(csv_path)
     _ensure_source_file(csv_path)
     payload = _build_local_payload(request_data, csv_path)
@@ -450,6 +460,11 @@ def run_single_scenario(scenario_name: str, csv_path: Path) -> dict:
         result['error'] = str(exc)
         return result
 
+    # Use scenario-specific CSV if mapped, otherwise use provided csv_path
+    if scenario_name in SCENARIO_CSV_MAPPING:
+        csv_path = SCENARIO_CSV_MAPPING[scenario_name]
+        print(f'  Using scenario-specific CSV: {csv_path}')
+    
     csv_path = _resolve_csv_path(csv_path)
 
     # Upload CSV to MinIO

tests/sientia/test_model_serving.py

@@ -112,18 +112,16 @@ def test_search_runs_by_name_with_order_by(mock_search_runs, mock_set_tracking_u
 def test_search_runs_by_name_raises_exception(
     mock_logging_error, mock_search_runs, mock_set_tracking_uri
 ):
-    """Test search_runs_by_name raises TypeError due to bug in line 80 of model_serving.py."""
+    """Test search_runs_by_name properly propagates SientiaMlException."""
     model_serving = ModelServing(tracking_uri='http://mlflow.example.com')
 
     exception = SientiaMlException(message='Search failed')
     mock_search_runs.side_effect = exception
 
-    # The code has a bug on line 80: "raise SientiaMlException from e"
-    # This raises TypeError because SientiaMlException requires 'message' argument
-    with raises(TypeError, match="missing 1 required positional argument: 'message'"):
+    with raises(SientiaMlException, match='Search failed'):
         model_serving.search_runs_by_name(['experiment1'])
 
-    mock_logging_error.assert_called_once()
+    mock_logging_error.assert_called_once_with(exception)
 
 
 @patch('model_manager.sientia.model_serving.mlflow.set_tracking_uri')

tests/utils/repository/test_training_repository.py

@@ -993,7 +993,9 @@ class TestConfigureDatetimeIndex:
         assert 'var2' in result.columns
 
     @pytest.mark.filterwarnings('ignore::UserWarning')
-    def test_configure_datetime_index_invalid_timestamp_column(self, training_repo, datetime_params):
+    def test_configure_datetime_index_invalid_timestamp_column(
+        self, training_repo, datetime_params
+    ):
         """Test _configure_datetime_index with invalid timestamp values."""
         data = pd.DataFrame(
             {
@@ -1199,10 +1201,12 @@ class TestEnsureDateColumnParsed:
 
     def test_parses_column_with_format(self, date_params):
         """When date_column and date_format set, column is parsed as datetime."""
-        data = pd.DataFrame({
+        data = pd.DataFrame(
+            {
                 'ts': ['2023-01-01 10:00:00', '2023-06-15 14:30:00'],
                 'x': [1, 2],
-        })
+            }
+        )
         result = _ensure_date_column_parsed(data, date_params)
         assert result['ts'].dtype == 'datetime64[ns]'
         assert result['ts'].iloc[0].year == 2023
@@ -1212,10 +1216,12 @@ class TestEnsureDateColumnParsed:
     def test_invalid_values_coerced_to_nat(self, date_params):
         """Invalid date strings are coerced to NaT when format is set."""
         date_params.date_format = 'yyyy-MM-dd HH:mm:ss'
-        data = pd.DataFrame({
+        data = pd.DataFrame(
+            {
                 'ts': ['2023-01-01 00:00:00', 'not-a-date', '2023-12-31 00:00:00'],
                 'x': [1, 2, 3],
-        })
+            }
+        )
         result = _ensure_date_column_parsed(data, date_params)
         assert pd.isna(result['ts'].iloc[1])
         assert result['ts'].iloc[0].year == 2023
@@ -1256,10 +1262,12 @@ class TestApplySupportFilters:
 
     def test_snake_case_upper_lower_line(self):
         """Support filters with upper_line/lower_line (snake_case) filter rows."""
-        data = pd.DataFrame({
+        data = pd.DataFrame(
+            {
                 'x': [1.0, 2.0, 3.0, 4.0],
                 'target': [2.0, 4.0, 6.0, 8.0],
-        })
+            }
+        )
         support_filters = {
             'x': {
                 'upper_line': {'intercept': 1.0, 'angle': 50},
@@ -1272,10 +1280,12 @@ class TestApplySupportFilters:
 
     def test_camel_case_upper_lower_line(self):
         """Support filters with upperLine/lowerLine (camelCase) are accepted."""
-        data = pd.DataFrame({
+        data = pd.DataFrame(
+            {
                 'x': [1.0, 2.0, 3.0],
                 'target': [1.0, 2.0, 3.0],
-        })
+            }
+        )
         support_filters = {
             'x': {
                 'upperLine': {'intercept': 2, 'angle': 5},
@@ -1288,11 +1298,13 @@ class TestApplySupportFilters:
 
     def test_two_variables_ands_masks(self):
         """Two variables apply AND of both masks."""
-        data = pd.DataFrame({
+        data = pd.DataFrame(
+            {
                 'a': [1.0, 2.0, 3.0],
                 'b': [1.0, 2.0, 3.0],
                 'target': [2.0, 2.0, 2.0],
-        })
+            }
+        )
         support_filters = {
             'a': {
                 'upper_line': {'intercept': 10, 'angle': 45},