PCA手写数字降维

import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import load_digits
from sklearn.neural_network import MLPClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report,confusion_matrix
from mpl_toolkits.mplot3d import Axes3D

def get_data():
    digits =  load_digits()
    x_data= digits.data
    y_data= digits.target
    return x_data,y_data

def zeroMean(dataMat):
    meanVal = np.mean(dataMat,axis = 0)
    newData = dataMat - meanVal
    return newData,meanVal
def draw(data):
    x_data= data[:,0]
    y_data= data[:,1]
    plt.scatter(x_data,y_data)
    plt.show()

def pca(dataMat,top):#将n维降为top维
    newData,meanVal = zeroMean(dataMat)
    covMat = np.cov(newData,rowvar=0)#求协方差矩阵
    eigVals,eigVects = np.linalg.eig(np.mat(covMat))#求特征值，特征向量
    eigValsIndice = np.argsort(eigVals)#对特征值排序，并显示下表顺序
    n_eigValsIndice = eigValsIndice[-1:-(top+1):-1]
    n_eigVect = eigVects[:,n_eigValsIndice ]
    lowDDataMat = newData*n_eigVect
    reconMat = (lowDDataMat*n_eigVect.T) + meanVal
    return (lowDDataMat,reconMat)
    
def main():
    print ("----------ing-------------")
    x_data,y_data = get_data()

    #print (x_data.shape)
    x_train,x_test,y_train,y_test = train_test_split(x_data,y_data)
    mlp = MLPClassifier(hidden_layer_sizes=(100,50),max_iter=500)
    mlp.fit(x_train,y_train)


    #print (lowDDataMat.shape)
    #plt.scatter(x,y,c='r')
    #plt.show()

    
    lowDDataMat,reconMat = pca(x_data,2)
    x = np.array(lowDDataMat)[:,0]
    y = np.array(lowDDataMat)[:,1]
    predictions = mlp.predict(x_data)
    plt.scatter(x,y,c=predictions)
    plt.show()

    lowDDataMat,reconMat = pca(x_data,3)
    x = np.array(lowDDataMat)[:,0]
    y = np.array(lowDDataMat)[:,1]
    z = np.array(lowDDataMat)[:,2]
    predictions = mlp.predict(x_data)
    ax = plt.figure().add_subplot(111,projection='3d')
    ax.scatter(x,y,z,c=predictions,s=10)
    plt.show()
    print ("----------end-------------")

main()