python主成分分析PCA完整代码以及结果图片

import numpy as np from pylab import * import matplotlib.pyplot as plt dataMat = [] x_axis_data = [ 15,16,20,20,35,38,42,44,45,45,50,56,57,62,65,70,80,85,55,65,78,80,87,97 ] y_axis_data = [ 30,32,36,37,77,22,79,25,36,62,18,46,25,39,68,64,60,52,88,87,85,84,77,60 ] for i in range(0,len(x_axis_data)): dataMat.append([x_axis_data[i],y_axis_data[i]]) print(dataMat)#原矩阵 k=1 # 设计pca函数 average = np.mean(dataMat, axis=0) #按列求均值 # print(average) m, n = np.shape(dataMat) meanRemoved = dataMat - np.tile(average, (m,1)) #减去均值 normData = meanRemoved / np.std(dataMat) #标准差归一化 # print(normData) average = np.mean(normData, axis=0) covMat = np.cov(normData.T) #求协方差矩阵 # covMat = np.cov(meanRemoved.T) # print(covMat) eigValue, eigVec = np.linalg.eig(covMat) #求协方差矩阵的特征值和特征向量 eigValInd = np.argsort(-eigValue) #返回特征值由大到小排序的下标 selectVec = np.matrix(eigVec.T[:k]) #因为[:k]表示前k行，因此之前需要转置处理(选择前k个大的特征值) # print(covMat) finalData = normData * selectVec.T #再转置回来 print(finalData)#pca之后的矩阵