多元线性回归代码

import pandas as pd
import statsmodels.api as sm
from sklearn.metrics import r2_score, mean_squared_error
import numpy as np
pd.set_option('display.max_rows', None)#显示全部行
from sklearn.model_selection import train_test_split
# 从Excel文件读取数据
# 假设Excel文件中有多列数据，包括植被指数（X1, X2, ...），最后一列是生物量（y）
# 请根据你的实际数据文件进行替换D:\桌面\修改生物量后建模-原始植被指数\2021.xls
data = pd.read_excel(r'D:   );#数据
 

# 定义自变量和因变量
X = data[['RVI' ]]  # 根据你的数据集修改列名 
y = data['Biomass']

# 将数据集分为训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)


# 从Excel文件读取数据
# 假设Excel文件中有多列数据，包括植被指数（X1, X2, ...），最后一列是生物量（y）
# 请根据你的实际数据文件进行替换

# 添加截距列
X_train = sm.add_constant(X_train)
X_test = sm.add_constant(X_test)

# 初始化多元线性回归模型
model = sm.OLS(y_train, X_train)

# 拟合模型
results = model.fit()

print(results.summary())

# 输出模型参数
coefficients = results.params
intercept = coefficients[0]

# 输出拟合方程
equation = f"y = {intercept:.4f} + "
for i, coef in enumerate(coefficients[1:]):
    equation += f"{coef:.4f} * X{i+1} + "
equation = equation.rstrip(" + ")

print("Fitted Equation:", equation)

# 在全部数据上进行预测
y_train_pred = results.predict(X_train)#在训练集预测
y_test_pred = results.predict(X_test)#在验证集预测


# 计算R方 训练集
r2_train = r2_score(y_train, y_train_pred)
# 计算RMSE和nRMSE
rmse_train = np.sqrt(mean_squared_error(y_train, y_train_pred))
nrmse_train = rmse_train / np.mean(y_train)

# 随机 计算R方 验证集
r2_test = r2_score(y_test, y_test_pred)
# 计算RMSE和nRMSE
rmse_test = np.sqrt(mean_squared_error(y_test, y_test_pred))
nrmse_test = rmse_test / np.mean(y_test)

# 输出结果

print("y训练集真实值",y_train)
print("y_pred训练集预测值:",y_train_pred)
print("y验证集真实值",y_test)
print("y_pred验证集预测值:",y_test_pred)

print("训练集R-squared:", r2_train)
print("训练集RMSE:", rmse_train)
print("训练集nRMSE:", nrmse_train)

print("验证集R-squared:", r2_test)
print("验证集RMSE:", rmse_test)
print("验证集nRMSE:", nrmse_test)