PCA

#学习自机器学习实战
from numpy import *
def loadDataSet(fileName,delim='\t'):
    fr=open(fileName)
    dataMat=[]
    for line in fr.readlines():
        curline=line.strip().split(delim)
        #print(curline)
        dataarr=[]
        n=len(curline)
        for j in range(n):
            dataarr.append(float(curline[j]))
        dataMat.append(dataarr)
    return dataMat
    # stringArr=[line.strip().split(delim) for line in fr.readlines()]
    # datArr=[map(float,line) for line in stringArr]
    # return mat(datArr)
def pca(dataMat,topNfeat=999999):
    meanVals=mean(dataMat,axis=0)#meanval
    meanRemoved=dataMat-meanVals#decrease meanval
    covMat=cov(meanRemoved,rowvar=0)#covariance
    eigVals,eigVects=linalg.eig(mat(covMat))#characteristic value and vector
    print("eigval:" ,eigVals)
    print("eigVects" ,eigVects)
    eigValInd=argsort(eigVals)#sort and list the serial number
    print("befor" ,eigValInd)
    eigValInd=eigValInd[:-(topNfeat+1):-1]#reverse the eigValInd
    print("after:")
    print(eigValInd)
    redEigVects=eigVects[:,eigValInd]
    print("redEigVects" ,redEigVects)
    lowDateMat=meanRemoved*redEigVects
    reconMat=(lowDateMat*redEigVects.T)+meanVals#restore
    return lowDateMat,reconMat
if __name__=='__main__':
    x=loadDataSet('cdata.txt')
    print(pca(x))