Enhance test coverage

Author: thomasmeissnercrm

bluecast/tests/conftest.py

@@ -3,8 +3,6 @@
 Session-scoped fixtures avoid redundant model training across test files.
 """
 
-import numpy as np
-import pandas as pd
 import pytest
 
 from bluecast.blueprints.cast import BlueCast
@@ -27,12 +25,6 @@
 )
 
 
-@pytest.fixture(scope="session")
-def fast_training_config():
-    """TrainingConfig tuned for speed in tests."""
-    return TrainingConfig(**_FAST_TRAINING_CONFIG_KWARGS)
-
-
 @pytest.fixture(scope="session")
 def synthetic_binary_data():
     """200-row synthetic binary classification dataset."""
@@ -63,48 +55,3 @@ def trained_bluecast_regression(synthetic_regression_data):
     df = synthetic_regression_data.copy()
     automl.fit(df, target_col="target")
     return automl
-
-
-@pytest.fixture(scope="session")
-def small_inline_data():
-    """Standard 6-feature, 20-row DataFrame used across many tests."""
-    x_train = pd.DataFrame(
-        {
-            "feature1": [i for i in range(20)],
-            "feature2": [i for i in range(20)],
-            "feature3": [i for i in range(20)],
-            "feature4": [i for i in range(20)],
-            "feature5": [i for i in range(20)],
-            "feature6": [i for i in range(20)],
-        }
-    )
-    x_test = pd.DataFrame(
-        {
-            "feature1": [i for i in range(20)],
-            "feature2": [i for i in range(20)],
-            "feature3": [i for i in range(20)],
-            "feature4": [i for i in range(20)],
-            "feature5": [i for i in range(20)],
-            "feature6": [i for i in range(20)],
-        }
-    )
-    y_train = pd.Series([0] * 10 + [1] * 10)
-    y_test = pd.Series([0] * 10 + [1] * 10)
-    return x_train, x_test, y_train, y_test
-
-
-@pytest.fixture(scope="session")
-def small_regression_data():
-    """Standard 6-feature, 20-row DataFrame for regression tests."""
-    x = pd.DataFrame(
-        {
-            "feature1": list(range(20)),
-            "feature2": list(range(20)),
-            "feature3": list(range(20)),
-            "feature4": list(range(20)),
-            "feature5": list(range(20)),
-            "feature6": list(range(20)),
-        }
-    )
-    y = pd.Series(np.random.default_rng(42).standard_normal(20))
-    return x.copy(), x.copy(), y.copy(), y.copy()

bluecast/tests/test_coverage_gaps.py

@@ -0,0 +1,97 @@
+"""Targeted tests to cover previously uncovered code paths."""
+
+import pandas as pd
+
+from bluecast.config.training_config import (
+    CatboostFinalParamConfig,
+    CatboostRegressionFinalParamConfig,
+    CatboostTuneParamsConfig,
+    CatboostTuneParamsRegressionConfig,
+    TrainingConfig,
+    XgboostFinalParamConfig,
+    XgboostRegressionFinalParamConfig,
+    XgboostTuneParamsConfig,
+    XgboostTuneParamsRegressionConfig,
+)
+from bluecast.preprocessing.train_test_split import train_test_split_time
+
+
+class TestConfigRepr:
+    def test_training_config_repr(self):
+        tc = TrainingConfig()
+        r = repr(tc)
+        assert "TrainingConfig(" in r
+        assert "global_random_state=33" in r
+
+    def test_xgboost_tune_params_repr(self):
+        cfg = XgboostTuneParamsConfig()
+        r = repr(cfg)
+        assert "XgboostTuneParamsConfig(" in r
+        assert "max_depth_min" in r
+
+    def test_xgboost_tune_params_regression_repr(self):
+        cfg = XgboostTuneParamsRegressionConfig()
+        r = repr(cfg)
+        assert "XgboostTuneParamsRegressionConfig(" in r
+
+    def test_xgboost_final_param_repr(self):
+        cfg = XgboostFinalParamConfig()
+        r = repr(cfg)
+        assert "XgboostFinalParamConfig(" in r
+        assert "params=" in r
+
+    def test_xgboost_regression_final_param_repr(self):
+        cfg = XgboostRegressionFinalParamConfig()
+        r = repr(cfg)
+        assert "XgboostRegressionFinalParamConfig(" in r
+
+    def test_catboost_tune_params_repr(self):
+        cfg = CatboostTuneParamsConfig()
+        r = repr(cfg)
+        assert "CatboostTuneParamsConfig(" in r
+
+    def test_catboost_tune_params_regression_repr(self):
+        cfg = CatboostTuneParamsRegressionConfig()
+        r = repr(cfg)
+        assert "CatboostTuneParamsRegressionConfig(" in r
+
+    def test_catboost_final_param_repr(self):
+        cfg = CatboostFinalParamConfig()
+        r = repr(cfg)
+        assert "CatboostFinalParamConfig(" in r
+        assert "params=" in r
+
+    def test_catboost_regression_final_param_repr(self):
+        cfg = CatboostRegressionFinalParamConfig()
+        r = repr(cfg)
+        assert "CatboostRegressionFinalParamConfig(" in r
+
+
+class TestShapWaterfallIndicesElseBranch:
+    def test_shap_waterfall_indices_with_value(self):
+        tc = TrainingConfig(shap_waterfall_indices=[0, 5, 99])
+        assert tc.shap_waterfall_indices == [0, 5, 99]
+
+    def test_shap_waterfall_indices_default(self):
+        tc = TrainingConfig()
+        assert tc.shap_waterfall_indices == []
+
+
+class TestTrainTestSplitTimeFString:
+    def test_time_split_with_column(self):
+        df = pd.DataFrame(
+            {"a": range(50), "b": range(50), "order": range(50), "target": range(50)}
+        )
+        x_train, x_test, y_train, y_test = train_test_split_time(
+            df, "target", "order", 0.8
+        )
+        assert len(x_train) == 40
+        assert len(x_test) == 10
+
+    def test_time_split_without_column(self):
+        df = pd.DataFrame({"a": range(50), "b": range(50), "target": range(50)})
+        x_train, x_test, y_train, y_test = train_test_split_time(
+            df, "target", "", 0.8
+        )
+        assert len(x_train) == 40
+        assert len(x_test) == 10