PCA主成分分析处理鸢尾花数据(python)

完整代码：

import numpy as np
import pandas as pd


data = pd.read_csv("C:\\Users\\Administrator\\Desktop\\iris.csv",header=None) 
#data=np.array(data)
data=data[[0,1,2,3]]
mean_data=np.mean(data,axis=0)
meanRemoved = data - mean_data   #去中心
covdata = np.cov(meanRemoved, rowvar=0) #求协方差矩阵
eigVals,eigVets=np.linalg.eig(np.mat(covdata)) #求解特征值和特征矩阵
eigValInd = np.argsort(eigVals)     #排序
eigValInd = eigValInd[:-(3+1):-1]   #保留最大的前2个特征值
redEigVects = eigVets[:,eigValInd]        #获得对应的特征向量
lowDData = np.mat(meanRemoved) * redEigVects   #获得获得降维后的特征

print("原数据集",meanRemoved)
print("降维后的数据集",lowDData)
print("累计方差贡献率为95%")
tot=sum(eigVals)
varExp=[(redEigVects/tot)*95 for redEigVects in sorted(eigVals,reverse=True)]
print(varExp)
cum_var_exp=np.cumsum(varExp)
print(cum_var_exp)