#k-means
#coding?utf-8
from?numpy?import?*
import?time
import?matplotlib.pyplot?as?plt

#calculate?Euclidean
def?euclDistance（vector1?vector2）:
????return?sqrt（sum（power（vector2?-?vector1?2）））

#init?centroids?with?random?sample
def?initCentroids（dataSet?k）:
????row?col?=?dataSet.shape
????centroids?=?zeros（（k?col））
????for?i?in?range（k）:
????????index?=?int（random.uniform（0?row））
????????centroids[i?:]?=?dataSet[index?:]
????return?centroids

#k-means?cluster
def?kmeans（dataSet?k）:
????row?col?=?dataSet.shape
????#first?column?stores?which?cluster?this?sample?belongs?to
????#sencond?column?stores?the?error?between?this?sample?and?its?centroid
????clusterAssment?=?mat（zeros（（row?2）））
????clusterChanged?=?True

????#?step1:?init?centroids
????centroids?=?initCentroids（dataSet?k）

????while?clusterChanged:
????????clusterChanged?=?False
????????#foreach?each?sample
????????for?i?in?xrange（row）:
????????????minDistance?=?1000
????????????minIndex?=?0
????????????#for?each?centroid
????????????##?step2:?find?the?centroid?who?is?the?closest
????????????for?j?in?range（k）:
????????????????distance?=?euclDistance（centroids[j:]?dataSet[i:]）
????????????????if?distance?????????????????????minDistance?=?distance
????????????????????minIndex?=?j

????????????##step3:update?its?cluster
????????????if?clusterAssment[i0]?!=?minIndex:
????????????????clusterChanged?=?True
????????????????clusterAssment[i:]?=?minIndex?minDistance**2

????????#step4?update?its?centroids
????????for?j?in?range（k）:
????????????pointsInCluster?=?dataSet[nonzero（clusterAssment[:?0].A?==?j）[0]]
????????????centroids[j:]?=?mean（pointsInCluster?axis?=?0）

????print?‘Congratulations?k-means?cluster?is?completed!!!!!!!!!!!!!!!‘
????return?centroids?clusterAssment


#show?your?cluster?only?avaliable?with?2D?data
def?showCluster（dataSet?k?centroids?clusterAssment）:
????row?col?=?dataSet.shape
????if?col?!=?2:
????????print?“Sorry!?I?can?not?draw?because?the?dimension?of?your?data?is??not?2“
????????return?1

????mark?=?[‘or‘?‘ob‘?‘og‘?‘ok‘?‘^r‘?‘+r‘?‘sr‘?‘dr‘?‘????if（k?>?len（mark））:
????????print?“Sorry!?Your?k?is?too?large!?Please?choose?smaller?k“
????????return?1

????#draw?all?samples
????for?i?in?xrange（row）:
????????markIndex?=?int（clusterAssment[i0]）
????????plt.plot（dataSet[i?0]?dataSet[i1]?mark[markIndex]）

????mark?=?[‘Dr‘?‘Db‘?‘Dg‘?‘Dk‘?‘^b‘?‘+b‘?‘sb‘?‘db‘?‘????#draw?the?centroids
????for?j?in?range（k）:
????????plt.plot（centroids[j0]?centroids[j1]?mark[j]?markersize?=?12）

????plt.show（）


#write?centroids?and?clusterAssment?in?txt
def?writeResult（dataSet?k?centroids?clusterAssment）:
????fileOut?=?open（“C:\Users\Administrator\Desktop\kmeans-python\centroids.txt“?“w“）
????out?=?open（“C:\Users\Administrator\Desktop\kmeans-python\clusterAssment.txt“?“w“）
????for?i?in?range（k）:
????????fileOut.write（str（c

?屬性????????????大小?????日期????時間???名稱
-----------?---------??----------?-----??----

?????文件?????????93??2016-12-06?17:01??kmeans-python\centroids.txt

?????文件???????3229??2016-12-06?17:01??kmeans-python\clusterAssment.txt

?????文件????????237??2016-12-06?15:41??kmeans-python\data.txt

?????文件???????1930??2016-12-06?15:40??kmeans-python\data2.txt

?????文件?????????27??2016-12-06?16:35??kmeans-python\haha.txt

?????文件???????3283??2016-12-06?17:01??kmeans-python\kmeans.py

?????文件???????3393??2016-12-06?17:01??kmeans-python\kmeans.pyc

?????文件????????234??2016-12-06?16:35??kmeans-python\test.py

?????文件????????750??2016-12-06?16:47??kmeans-python\test_kmeans.py

?????目錄??????????0??2016-12-06?17:01??kmeans-python

-----------?---------??----------?-----??----

????????????????13176????????????????????10