k-近邻分类的Python实现

参见《机器学习实战》

 1 # -*- coding:cp936 -*-

 2 #===============================================================================

 3 # 设计KNN最近邻分类器：

 4 #     找出每个元素在数据集中的最近邻的K个数据，统计这K个数据所属的类，所属类最多的那个类就是该元素所属的类

 5 #===============================================================================

 6 import numpy as np

 7 

 8 def loadHaiLunData(f_name):

 9     with open(f_name) as fHandle:

10         fLines = fHandle.readlines()

11         dataLines = len(fLines)

12         label  = []

13         dataSetMat = np.zeros((dataLines,3))

14         for i in range(dataLines):

15             lineList = fLines[i].strip().split('\t')

16             dataSetMat[i,:] = lineList[0:3]

17             label.append(int(lineList[-1]))

18         return dataSetMat,label

19     

20 

21 def dataNorm(dataSet):

22     numOfEle = dataSet.shape[0]

23     minEle = dataSet.min(0)

24     maxEle = dataSet.max(0)

25     normedData = (dataSet-np.tile(minEle,(numOfEle,1)))/np.tile(maxEle-minEle,(numOfEle,1))

26     return normedData

27 

28 def classifyKnn(inX, dataSet, label, k):

29     #===========================================================================

30     # inX：输入向量

31     # dataSet:保存数据特征的数组，每一行为若干个特征的参数，与label对应

32     # label：表明当前这个数据集中的每一个元素属于哪一类

33     # k：设定最近邻的个数

34     #===========================================================================

35     

36     #首先对数据集进行归一化

37 #     dataSet = dataNorm(dataSet)

38     numOfEle = dataSet.shape[0]

39     index = 0

40     diffDistance = dataSet - np.tile(inX, (numOfEle,1))

41     diffDistance = diffDistance**2

42     squareDistance = diffDistance.sum(1)

43 #     squareDistance = squareDistance**0.5

44     knnIndex = squareDistance.argsort()

45     #统计最近的k个近邻的label，看哪个label类别最多就可将该训练元素判为对应类

46     staticDict = {}

47     for i in range(k):

48         staticDict[label[knnIndex[i]]]=staticDict.get(label[knnIndex[i]],0)+1

49     itemList = staticDict.items()

50     argmax = np.argmax(itemList, axis = 0)

51     return itemList[argmax[1]][0]

52     

53 def testHaiLunClassify(k = 3, hRatio = 0.5):

54     dataSet,label = loadHaiLunData('datingTestSet2.txt')

55 #     hRatio = 0.5

56     totalNum = dataSet.shape[0]

57     testNum = int(totalNum*hRatio)

58     dataNormed = dataNorm(dataSet)

59     errorClass = 0

60     for i in range(testNum):

61         classRes = classifyKnn(dataNormed[i,:], dataNormed[testNum:,:], label[testNum:], k)

62         if classRes != label[i]:

63             errorClass += 1

64 #             print "classify error, No. %d should be label %d but got %d"%(i, label[i],classRes)

65     errorRate = errorClass/float(testNum)

66 #     print

67 #     print "Error rate: %f"%(errorRate)

68     return errorRate

69 

70 if __name__ == '__main__':

71     errorList = []

72     kRange = range(1,50,1)

73     for k in kRange:

74         errorList.append(testHaiLunClassify(k))

75     print errorList

76     import matplotlib.pyplot as  plt

77     fig = plt.figure(1)

78 #     ax  = fig.add_subplot(111)

79     plt.plot(kRange, errorList,'rs-')

80     plt.show()

81