python--knn算法练习

1，准备数据：

           从文件中读出文件（数据来自海伦约会网站例子）；展示如图 

                      

  

2 ，数据归一化

 主要公式：newvalue =（oldvalue-min）/（max-min）

>>> tmpmat,rangs,mins =knn.autonorm(mat)
>>> tmpmat
array([[ 0.44832535,  0.39805139,  0.56233353],
       [ 0.15873259,  0.34195467,  0.98724416],
       [ 0.28542943,  0.06892523,  0.47449629],
       ..., 
       [ 0.29115949,  0.50910294,  0.51079493],
       [ 0.52711097,  0.43665451,  0.4290048 ],
       [ 0.47940793,  0.3768091 ,  0.78571804]])

3，分类knn算法代码

def    classify(testvecs,dataset,labels,k):
    datasetsize =dataset.shape[0]
    diffmat =tile(testvecs,(datasetsize,1)) -dataset
    sqdiffmat= diffmat**2
    sqdistance=sqdiffmat.sum(axis=1)
    distance =sqdistance**0.5
    sorteddiss=distance.argsort()
    classcount = {}
    for i in range(k):
        votelabel =labels[sorteddiss[i]]
        classcount[votelabel] = classcount.get(votelabel,0)+1.0
    sortedclasscount =sorted(classcount.iteritems(),key=operator.itemgetter(1),reverse=True)

4，测试

def datingtest(int_k):
    horatio =0.1
    datingdatamat ,labels =file2Mat('D:\python-机器学习\datingTestSet2.txt')
    normmat,ranges,minvals =autonorm(datingdatamat)
    m=normmat.shape[0]
    numtestvecs =int (m*horatio)
    errorcount = 0.0
    for i in range (numtestvecs):
        classifyresult =classify(normmat[i,:],normmat[numtestvecs:m,:],labels[numtestvecs:m],int_k)
        if(classifyresult != labels[i]):
            errorcount=errorcount+1.0

    print "the toal error rate is :%f" %(errorcount/float(numtestvecs))

>>> reload(knn)
<module 'knn' from 'D:\python-机器学习\knn.py'>
>>> knn.datingtest(4)
the toal error rate is :0.030000

出错的点在交集处