KNN算法

根据机器学习实战上的代码，又自己完成KNN算法的简单实现，在这里做一个记录。
需要注意的问题是：
1.diatance是一个series，需要根据数值排序，
2.根据sortedDistIndicies，到y_train中寻找key时，需要用y_train[y_train.index==sortedDistIndicies.index[i]]。不能直接取y_train[sortedDistIndicies.index[i]]

filename=r'./datingTestSet.txt'
file=pd.read_csv(filename,sep='\t',header=None)
returnMat=file.iloc[:,0:3]
data_norm = returnMat.apply(lambda x: (x - np.min(x)) / (np.max(x)-np.min(x)))
data_label=file.iloc[:,-1]

x_train, x_test, y_train, y_test = train_test_split(data_norm, data_label, test_size = 0.1)

errorCount=0.0
for x,y in zip(np.array(x_test),y_test):
    dataSetSize = x_train.shape[0]
    diffMat = np.tile(x, (dataSetSize, 1)) - x_train
    sqDiffMat = diffMat ** 2
    sqDistances = sqDiffMat.sum(axis=1)
    distance = sqDistances ** 0.5
    sortedDistIndicies = distance.sort_values()
    classCount = {}
    for i in range(3):
        voteilabel = y_train[y_train.index==sortedDistIndicies.index[i]]
        voteIlabel=voteilabel.values[0]
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1
    sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True)
    classifierResult=sortedClassCount[0][0]
    # classifierResult=classify0(x,x_train,list(y_train),3)
    print('the classifier came back with:%s,the real answer is:%s'%(classifierResult,y))
    if (classifierResult != y):
        errorCount+=1
print('the total error rate is : %f'%(errorCount/float(len(x_test))))

knn_classifier=KNeighborsClassifier(3)
knn_classifier.fit(x_train,y_train)
y_predict=knn_classifier.predict(x_test)
scores=knn_classifier.score(x_test,y_test)
print('acc:{}'.format(scores))