Python聚類算法之基本K均值實例詳解

本文實例講述了Python聚類算法之基本K均值運算技巧。分享給大家供大家參考，具體如下：
基本K均值 ：選擇 K 個初始質(zhì)心，其中 K 是用戶指定的參數(shù)，即所期望的簇的個數(shù)。每次循環(huán)中，每個點被指派到最近的質(zhì)心，指派到同一個質(zhì)心的點集構(gòu)成一個。然后，根據(jù)指派到簇的點，更新每個簇的質(zhì)心。重復指派和更新操作，直到質(zhì)心不發(fā)生明顯的變化。    
# scoding=utf-8
import pylab as pl
points = [[int(eachpoint.split("#")[0]), int(eachpoint.split("#")[1])] for eachpoint in open("points","r")]
# 指定三個初始質(zhì)心
currentCenter1 = [20,190]; currentCenter2 = [120,90]; currentCenter3 = [170,140]
pl.plot([currentCenter1[0]], [currentCenter1[1]],'ok')
pl.plot([currentCenter2[0]], [currentCenter2[1]],'ok')
pl.plot([currentCenter3[0]], [currentCenter3[1]],'ok')
# 記錄每次迭代后每個簇的質(zhì)心的更新軌跡
center1 = [currentCenter1]; center2 = [currentCenter2]; center3 = [currentCenter3]
# 三個簇
group1 = []; group2 = []; group3 = []
for runtime in range(50):
  group1 = []; group2 = []; group3 = []
  for eachpoint in points:
    # 計算每個點到三個質(zhì)心的距離
    distance1 = pow(abs(eachpoint[0]-currentCenter1[0]),2) + pow(abs(eachpoint[1]-currentCenter1[1]),2)
    distance2 = pow(abs(eachpoint[0]-currentCenter2[0]),2) + pow(abs(eachpoint[1]-currentCenter2[1]),2)
    distance3 = pow(abs(eachpoint[0]-currentCenter3[0]),2) + pow(abs(eachpoint[1]-currentCenter3[1]),2)
    # 將該點指派到離它最近的質(zhì)心所在的簇
    mindis = min(distance1,distance2,distance3)
    if(mindis == distance1):
      group1.append(eachpoint)
    elif(mindis == distance2):
      group2.append(eachpoint)
    else:
      group3.append(eachpoint)
  # 指派完所有的點后，更新每個簇的質(zhì)心
  currentCenter1 = [sum([eachpoint[0] for eachpoint in group1])/len(group1),sum([eachpoint[1] for eachpoint in group1])/len(group1)]
  currentCenter2 = [sum([eachpoint[0] for eachpoint in group2])/len(group2),sum([eachpoint[1] for eachpoint in group2])/len(group2)]
  currentCenter3 = [sum([eachpoint[0] for eachpoint in group3])/len(group3),sum([eachpoint[1] for eachpoint in group3])/len(group3)]
  # 記錄該次對質(zhì)心的更新
  center1.append(currentCenter1)
  center2.append(currentCenter2)
  center3.append(currentCenter3)
# 打印所有的點，用顏色標識該點所屬的簇
pl.plot([eachpoint[0] for eachpoint in group1], [eachpoint[1] for eachpoint in group1], 'or')
pl.plot([eachpoint[0] for eachpoint in group2], [eachpoint[1] for eachpoint in group2], 'oy')
pl.plot([eachpoint[0] for eachpoint in group3], [eachpoint[1] for eachpoint in group3], 'og')
# 打印每個簇的質(zhì)心的更新軌跡
for center in [center1,center2,center3]:
  pl.plot([eachcenter[0] for eachcenter in center], [eachcenter[1] for eachcenter in center],'k')

pl.show()

運行效果截圖如下：

希望本文所述對大家Python程序設計有所幫助。