python数据预处理 :数据共线性处理详解

import numpy as np
import pandas as pd
from sklearn.linear_model import Ridge
from sklearn.decomposition import PCA
from sklearn.linear_model import LinearRegression

# 导入数据
df = pd.read_csv('https://raw.githubusercontent.com/ffzs/dataset/master/boston/train.csv')

# 切分自变量
X = df.iloc[:, 1:-1].values

# 切分预测变量
y = df.iloc[:, [-1]].values

# 使用岭回归处理
import matplotlib.pyplot as plt
plt.figure(figsize=(8,6))
n_alphas = 20
alphas = np.logspace(-1,4,num=n_alphas)
coefs = []
for a in alphas:
  ridge = Ridge(alpha=a, fit_intercept=False)
  ridge.fit(X, y)
  coefs.append(ridge.coef_[0])
ax = plt.gca()
ax.plot(alphas, coefs)
ax.set_xscale('log')
handles, labels = ax.get_legend_handles_labels()
plt.legend(labels=df.columns[1:-1])
plt.xlabel('alpha')
plt.ylabel('weights')
plt.axis('tight')
plt.show()

# 主成分回归进行回归分析
pca_model = PCA()
data_pca = pca_model.fit_transform(X)

# 得到所有主成分方差
ratio_cumsum = np.cumsum(pca_model.explained_variance_ratio_)
# 获取方差占比超过0.8的索引值
rule_index = np.where(ratio_cumsum > 0.9)
# 获取最小的索引值
min_index = rule_index[0][0]
# 根据最小索引值提取主成分
data_pca_result = data_pca[:, :min_index+1]
# 建立回归模型
model_liner = LinearRegression()
# 训练模型
model_liner.fit(data_pca_result, y)
print(model_liner.coef_)
#[[-0.02430516 -0.01404814]]