更新 function.py

function.py

@@ -65,8 +65,7 @@ def train_logreg(X, y, learning_rate=0.1, n_estimators=200, max_depth=5):
     - max_depth (int): Maximum depth of each tree.
 
     Returns:
-    - GradientBoostingClassifier: Trained Gradient Boosting model as weights.
-    - StandardScaler: Scaler fitted on training data for consistent scaling.
+    - tuple: (Trained model and scaler) packed into 'weights' for predict_logreg compatibility.
     """
     scaler = StandardScaler()
     X = scaler.fit_transform(X)
@@ -79,23 +78,23 @@ def train_logreg(X, y, learning_rate=0.1, n_estimators=200, max_depth=5):
     )
     gb_model.fit(X, y)
     
-    # Return model and scaler separately
-    return gb_model, scaler
+    # Pack both model and scaler as weights
+    return (gb_model, scaler)
 
 
-def predict_logreg(inX, model, scaler):
+def predict_logreg(inX, weights):
     """
     Predict the class label for a given sample using Gradient Boosting.
 
     Args:
     - inX (numpy array): Input sample for classification.
-    - model (GradientBoostingClassifier): Trained model.
-    - scaler (StandardScaler): Scaler for feature normalization.
+    - weights (tuple): Packed tuple containing trained model and scaler.
 
     Returns:
     - int: Predicted class label (0 or 1).
     """
-    inX = scaler.transform([inX])
+    model, scaler = weights  # Unpack model and scaler
+    inX = scaler.transform([inX])  # Scale input
     prediction = model.predict(inX)
     return int(prediction[0])
 
@@ -106,8 +105,8 @@ if __name__ == "__main__":
     X_test, y_test = load_data("./data/horse-colic.test")
 
     # 训练模型
-    model, scaler = train_logreg(X_train, y_train)
+    weights = train_logreg(X_train, y_train)
 
     # 测试单个数据
-    y_pred = predict_logreg(X_test[11], model, scaler)
+    y_pred = predict_logreg(X_test[11], weights)
     print(f"预测结果: {y_pred}, 实际结果: {int(y_test[11])}")