K-means使用python进行实现

from scipy.cluster.vq import *

import numpy as np

from matplotlib import pyplot as plt

class1=1.5*np.random.randn(100,2)

##print(class1)

class2=np.random.randn(100,2)+np.array([8,8])

##print(class2)

features=np.vstack((class1,class2))

centroids,variance=kmeans(features,2)

code,distance=vq(features,centroids)

plt.figure()

ndx=np.where(code==1)[0]

plt.plot(features[ndx,0],features[ndx,1],'*')

ndx=np.where(code==0)[0]

plt.plot(features[ndx,0],features[ndx,1],'r.')

plt.plot(centroids[:,0],centroids[:,1],'go')

plt.axis('off')

plt.show()