#判断适合不适合做回归分析========================================================================== # 导入库 import numpy as np import pandas as pd import matplotlib.pyplot as plt import sklearn from sklearn.linear_model import LinearRegression # 读取数据 data = pd.read_excel('data1.xlsx') X = data.iloc[:, :-1].values # 特征, 取所有行,但是不要最后一列,就是只取第一列 y = data.iloc[:, -1].values # 响应变量,取最后一列 # 绘制散点图 plt.scatter(X[:,0], y) plt.xlabel('X1') plt.ylabel('y') plt.show() # 计算相关系数 corr = np.corrcoef(X[:,0], y)[0, 1] print('相关系数为:', corr) # 拟合模型 model = LinearRegression() model.fit(X, y) # 输出R方值 R2 = model.score(X, y) print('R方值为:', R2) #预测======================================================================================= #导入库 import numpy as np import pandas as pd from sklearn.linear_model import LinearRegression from sklearn.metrics import mean_squared_error # 读取数据 data = pd.read_excel('data1.xlsx') X = data.iloc[:, :-1].values # 特征 y = data.iloc[:, -1].values # 响应变量 # 拟合模型 model = LinearRegression() model.fit(X, y) # 预测未来值 future_data = np.array([[63]]) future_pred = model.predict(future_data) # 计算置信度和置信区间 y_pred = model.predict(X) mse = mean_squared_error(y, y_pred) rmse = np.sqrt(mse) n = len(y) k = len(X[0]) alpha = 0.05 # 置信度水平 t_value = 2.064 # 自由度为n-k-1,置信度为95%时的t值 s = np.sqrt(mse / (n - k - 1)) se = s * np.sqrt(np.dot(np.dot(future_data, np.linalg.inv(np.dot(X.T, X))), future_data.T) + 1) confidence_interval = t_value * se lower_bound = future_pred - confidence_interval upper_bound = future_pred + confidence_interval # 输出结果 print('未来值预测为:', future_pred) print('置信区间为:', lower_bound, '-', upper_bound) print('置信度为95%的误差范围为:', rmse * t_value)
线性回归判断能否与预测实现python
于 2023-04-03 20:16:42 首次发布