聚類和決策樹一樣施蜜，屬于無監(jiān)督學(xué)習(xí)辈灼。也就是說數(shù)據(jù)樣本只有特征x，沒有給定y。聚類的目的是找到樣本特征潛在的類別翘单，將同類別的樣本放在一起。

kmeans的具體邏輯如下：
1.隨機(jī)選取k個(gè)簇心巾陕；
2.對(duì)于每一個(gè)樣例琳钉，計(jì)算其屬于的類；
3.循環(huán)完所有的樣例后戳表，重新計(jì)算每個(gè)簇的簇心桶至；
4.重復(fù)第二步第三部，直到簇心不再變化或達(dá)到最大迭代值匾旭。

import numpy as np
import matplotlib.pyplot as plt

#讀取數(shù)據(jù)
def loaddate(filename):
    datamat = []
    fr = open(filename)
    for line in fr.readlines():
        datamat.append(map(float, line.strip().split('\t')))
    datamat = pd.DataFrame(datamat)
    return datamat

#計(jì)算歐式距離
def distance(vecA, vecB):
    return np.sqrt(np.sum(np.power(vecA - vecB, 2)))

#產(chǎn)生隨機(jī)簇心
def getcenter(data, k):
    _, n = data.shape
    center = pd.DataFrame(np.zeros([k, 2]))
    for i in range(n):
        minJ = np.min(data.iloc[:, i])
        maxJ = np.max(data.iloc[:, i])
        rangJ = float(maxJ - minJ)
        center.iloc[:, i] = minJ +rangJ * np.random.rand(k, 1)
    return center

#計(jì)算每個(gè)樣例歸屬的簇镣屹，并重新計(jì)算簇心
def kmeans(data, k, maxiter):
    ceter = getcenter(data, k)
    m, n = data.shape
    position = pd.DataFrame(np.zeros((m, 2)), columns = ['dis', 'cindex'])
    #position第一列放置樣例距簇心的距離，第二列放置樣本歸屬的簇心
    itercount = 0
    #迭代次數(shù)价涝，如果迭代次數(shù)超過最大迭代次數(shù)女蜈，則停止
    clusterchange = True
    #簇心是否發(fā)生變化
    while itercount < maxiter and clusterchange:
        itercount += 1
        clusterchange = False
        for i in range(m):
            minindex = 0
            mindist = np.inf
            #設(shè)置position的初始值
            for j in range(k):
                dis = distance(data.iloc[i, :], ceter.iloc[j, :])
                #計(jì)算歐式距離
                if dis < mindist:
                    minindex = j
                    mindist = dis
                    #替換掉初始化的數(shù)據(jù)
            if position.iloc[i, 1] != minindex:
                clusterchange = True
                #觀察歸屬的簇是否發(fā)生變化
            position.iloc[i, :] = mindist, minindex
        for cent in range(k):
            ptscluster = data.iloc[list(position.loc[position['cindex'] == cent, 'cindex'].index), :]
            if ptscluster.shape[0] > 0:
                ceter.iloc[cent, :] = np.mean(ptscluster, axis = 0)
                #計(jì)算新的簇心
    return ceter
                

if __name__ == '__main__':
    file = loaddate('testSet.txt')
    ceter = kmeans(file, 2, 5)
    plt.scatter(file.iloc[:, 0], file.iloc[:, 1], marker = '*', c = 'b')
    plt.scatter(ceter.iloc[:, 0], ceter.iloc[:, 1], marker = 'o', c = 'r')
    plt.show()