from numpy import *
import operator
def createDataSet():
    group = array([[1.0, 1.1], [1.0, 1.0], [0, 0], [0, 0.1]])
    labels = ['A', 'A', 'B', 'B']
    return group, labels
# k-近邻算法
def classify0(inX, dataSet, labels, k):
    # 列出数组长度
    dataSetSize = dataSet.shape[0]
    # 算出离偏移量 的x，y轴距离
    diffMat = tile(inX, (dataSetSize, 1)) - dataSet
    # 对 x y 分别求平方
    sqDiffMat = diffMat ** 2
    # 将x y 求和
    sqDistances = sqDiffMat.sum(axis=1)
    # 开方算出距离 即 d = /x2 + y2
    distances = sqDistances ** 0.5
    # 返回距离从小到大的索引值
    sortedDistIndicies = distances.argsort()
    classCount = {}
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        # get方法返回classCount中键名对应的值 没有则返回默认值
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1
    # 根据键值进行降序排列
    sortedClassCount = sorted(classCount.items(),
                              key=operator.itemgetter(1), reverse=True)
    return sortedClassCount[0][0]
group, labels = createDataSet()
print(classify0((0, 0), group, labels, 3))