Pre-commit

Author: thomasmeissnercrm

bluecast/tests/test_ai_module.py

@@ -32,7 +32,10 @@ def test_defaults(self):
         assert config.checkpoint_dir is None
 
     def test_get_model_name_default(self):
-        assert AIConfig(api_key="t", provider="gemini").get_model_name() == "gemini-2.5-flash"
+        assert (
+            AIConfig(api_key="t", provider="gemini").get_model_name()
+            == "gemini-2.5-flash"
+        )
         assert AIConfig(api_key="t", provider="openai").get_model_name() == "gpt-4o"
         assert (
             AIConfig(api_key="t", provider="anthropic").get_model_name()
@@ -45,9 +48,13 @@ def test_get_model_name_custom(self):
 
     def test_get_max_iterations(self):
         assert AIConfig(api_key="t", max_iterations=7).get_max_iterations() == 7
-        assert AIConfig(api_key="t", max_iterations=0, mode="fast").get_max_iterations() == 1
         assert (
-            AIConfig(api_key="t", max_iterations=0, mode="precise").get_max_iterations() == 5
+            AIConfig(api_key="t", max_iterations=0, mode="fast").get_max_iterations()
+            == 1
+        )
+        assert (
+            AIConfig(api_key="t", max_iterations=0, mode="precise").get_max_iterations()
+            == 5
         )
 
 
@@ -161,7 +168,9 @@ def test_check_leakage(self, sample_df):
         assert "leakage" in result.lower()
 
     def test_create_feature_success(self, sample_df):
-        result = tool_create_feature(sample_df, "df['ratio'] = df['num1'] / (df['num2'] + 1)")
+        result = tool_create_feature(
+            sample_df, "df['ratio'] = df['num1'] / (df['num2'] + 1)"
+        )
         assert result["success"] is True
         assert "ratio" in result["new_columns"]
 

bluecast/tests/test_conformal_group.py

@@ -2,7 +2,6 @@
 
 import numpy as np
 import pandas as pd
-import pytest
 
 from bluecast.conformal_prediction.conformal_prediction import (
     ConformalPredictionWrapper,
@@ -139,9 +138,7 @@ def test_coverage_by_group(self):
         )
         groups = np.array(["A", "A", "A", "A", "B", "B", "B", "B"])
 
-        result = prediction_interval_coverage_by_group(
-            y_true, intervals, [0.1], groups
-        )
+        result = prediction_interval_coverage_by_group(y_true, intervals, [0.1], groups)
         assert "A" in result
         assert "B" in result
         assert 0.1 in result["A"]
@@ -167,15 +164,19 @@ def test_all_fair(self):
             "urban": {0.1: 0.91},
             "rural": {0.1: 0.89},
         }
-        result = conformal_fairness_check(coverages, target_coverage=0.9, tolerance=0.05)
+        result = conformal_fairness_check(
+            coverages, target_coverage=0.9, tolerance=0.05
+        )
         assert result["is_fair"] is True
 
     def test_unfair(self):
         coverages = {
             "urban": {0.1: 0.92},
             "rural": {0.1: 0.78},
         }
-        result = conformal_fairness_check(coverages, target_coverage=0.9, tolerance=0.05)
+        result = conformal_fairness_check(
+            coverages, target_coverage=0.9, tolerance=0.05
+        )
         assert result["is_fair"] is False
         assert result["coverage_range"] == (0.78, 0.92)
 

bluecast/tests/test_fairness.py

@@ -6,8 +6,6 @@
 
 from bluecast.evaluation.fairness import (
     FairnessAuditor,
-    FairnessReport,
-    GroupMetrics,
     _safe_ratio,
 )
 
@@ -140,9 +138,7 @@ def test_fairness_auditor_missing_column(binary_data):
 
 def test_fairness_auditor_reference_group(binary_data):
     y_true, y_pred, y_probs, df = binary_data
-    auditor = FairnessAuditor(
-        sensitive_columns=["gender"], reference_group="female"
-    )
+    auditor = FairnessAuditor(sensitive_columns=["gender"], reference_group="female")
     reports = auditor.audit_classification(y_true, y_pred, y_probs, df)
     report = reports[0]
     for key in report.demographic_parity:

bluecast/tests/test_unified.py

@@ -1,13 +1,11 @@
 """Tests for the unified BlueCastAuto interface."""
 
-import numpy as np
 import pandas as pd
 import pytest
 from sklearn.datasets import make_classification, make_regression
 
 from bluecast.blueprints.unified import BlueCastAuto
 from bluecast.config.training_config import TrainingConfig
-from bluecast.ensemble.ensemble_config import EnsembleConfig
 
 
 @pytest.fixture
@@ -66,7 +64,9 @@ def test_binary_cv(binary_data, fast_config):
 def test_regression_single_model(regression_data, fast_config):
     from sklearn.model_selection import train_test_split
 
-    df_train, df_eval = train_test_split(regression_data, test_size=0.2, random_state=42)
+    df_train, df_eval = train_test_split(
+        regression_data, test_size=0.2, random_state=42
+    )
     y_eval = df_eval.pop("target")
 
     automl = BlueCastAuto(