Update app.py

app.py

@@ -4,30 +4,32 @@ import joblib
 import numpy as np
 from sklearn.ensemble import RandomForestRegressor
 
-# 模拟加载或创建模型
-try:
-    model = joblib.load('rf_features.joblib')
-except:
-    # 模拟数据训练
-    X = pd.DataFrame(np.random.rand(100, 2), columns=['feature1', 'feature2'])
-    y = X['feature1'] * 3 + X['feature2'] * 2 + np.random.rand(100)
-    model = joblib.load("rf_features.joblib")
-    model.fit(X, y)
-    joblib.dump(model, 'model.pkl')
+# ——— 加载你训练好的模型和特征名 ———
+model = joblib.load("rf_model.joblib")
+feature_names = joblib.load("rf_features.joblib")  # 应当是像 ["feature1", "feature2", ...]
 
-# 预测函数
-def predict_energy(feature1, feature2):
-    input_data = pd.DataFrame([[feature1, feature2]], columns=['feature1', 'feature2'])
-    prediction = model.predict(input_data)
-    return f'预测能耗为：{prediction[0]:.2f}'
+# ——— 定义预测函数 ———
+def predict_energy(*inputs):
+    # 将输入打包成 DataFrame
+    df = pd.DataFrame([inputs], columns=feature_names)
+    # 调用模型预测
+    pred = model.predict(df)
+    return f"预测能耗为：{pred[0]:.2f}"
+
+# ——— 构建 Gradio 界面 ———
+# 动态为每个特征创建一个 Number 输入框
+input_components = [
+    gr.Number(label=name) for name in feature_names
+]
 
-# Gradio 接口定义
 demo = gr.Interface(
     fn=predict_energy,
-    inputs=[gr.Number(label='Feature 1'), gr.Number(label='Feature 2')],
-    outputs='text',
-    title='Energy Predictor',
-    description='根据输入特征预测能耗'
+    inputs=input_components,
+    outputs="text",
+    title="🏡 Energy Usage Predictor",
+    description="基于 RandomForest 已训练模型 (rf_model.joblib)，输入若干特征，输出能耗预测结果"
 )
 
-demo.launch()
+# ——— 启动 ———
+if __name__ == "__main__":
+    demo.launch()