import pandas as pd
from ucimlrepo import fetch_ucirepo
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
import joblib
import matplotlib.pyplot as plt


# 获取数据集
student_performance = fetch_ucirepo(id=320)

# 获取特征和目标
X = student_performance.data.features
y = student_performance.data.targets

# 查看特征和目标的前几行
print(X.head())
print(y.head())

# 编码分类变量
X = pd.get_dummies(X, drop_first=True)

# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y['G3'], test_size=0.2, random_state=42)

# 创建并训练模型
model = RandomForestRegressor(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

# 保存模型
model_path = "C:/Users/baby7/Desktop/推理/model_checkpoints/random_forest_model.pkl"
joblib.dump(model, model_path)
print(f"模型已保存到 {model_path}")

# 加载模型
loaded_model = joblib.load(model_path)
print("模型已加载")

# 使用加载的模型进行预测
y_pred = loaded_model.predict(X_test)  # X_test 是您的测试数据
print("预测结果:", y_pred)

# 评估模型性能
from sklearn.metrics import mean_squared_error

mse = mean_squared_error(y_test, y_pred)
print(f'均方误差: {mse:.2f}')

import matplotlib.pyplot as plt

plt.scatter(y_test, y_pred)
plt.xlabel('真实值')
plt.ylabel('预测值')
plt.title('真实值与预测值对比')
plt.plot([0, 20], [0, 20], color='red', linestyle='--')  # 参考线
plt.show()