script for testing bad features

scripts/bad_features_test.py

@@ -5,7 +5,7 @@
 
 # Original dataset
 X_original = pd.DataFrame(np.random.rand(1000, 10))  # 10 original features
-y = (X_original.sum(axis=1) > 5).astype(int)         # Simple binary target
+y = (X_original.sum(axis=1) > 5).astype(int)  # Simple binary target
 
 # Add random noise features
 X_noise = X_original.copy()
@@ -18,8 +18,8 @@
 # Train and compare models
 model = XGBClassifier()
 
-original_score = cross_val_score(model, X_original, y, cv=5, scoring='accuracy').mean()
-noise_score = cross_val_score(model, X_noise, y, cv=5, scoring='accuracy').mean()
+original_score = cross_val_score(model, X_original, y, cv=5, scoring="accuracy").mean()
+noise_score = cross_val_score(model, X_noise, y, cv=5, scoring="accuracy").mean()
 
 print(f"Original CV Accuracy: {original_score:.4f}")
 print(f"CV Accuracy with Adversarial Features: {noise_score:.4f}")