k近邻法

 1 #-*-coding:utf-8 -*-
 2 import numpy as np
 3 import math
 4 
 5 #训练集和类别
 6 def creatDataSet( ):
 7     group = np.array([[1.0,2.0],[1.2,0.1],[0.1,1.4],[0.3,3.5]])
 8     label = [‘A‘,‘A‘,‘B‘,‘B‘]
 9     return group,label
10 
11 #确定输入实例以及k值
12 def inputData( ):
13     test_data=np.array([ [1.1 , 0.3] ])
14     K = 3
15     return test_data,K
16 
17 #归一化
18 def Normalization( group ):
19     max = np.amax(group,axis=0)
20     min = np.amin(group,axis=0)
21     for i in range(len(group)):
22         group[i][0] /= (max[ 0 ] - min[ 0 ])
23         group[i][1] /= (max[ 1 ] - min[ 1 ])
24     return group
25 
26 #求输入实例与训练集中所有点的欧氏距离
27 def Distance( group ,dis ):
28     test,K=inputData( )
29     group = Normalization( group )
30     for i in range(len(dis)):
31         dis[i] = math.sqrt((group[i][0] - test[0][0]) ** 2 + (group[i][1] - test[0][1]) ** 2 )
32     return dis,K,test
33 
34 #找出输入实例的类别
35 def Classify_By_KNN( group , label ):
36     size = len( label )
37     dis = np.zeros( size )
38     dis,k,test = Distance( group ,dis )
39     sortedDistIndex=np.argsort(dis)
40     countLabel={}
41     for i in range(k):
42         l=label[sortedDistIndex[i]]
43         countLabel[l] = countLabel.get(l,0) + 1
44     countMax=max(countLabel.values())
45     for key,value in countLabel.items():
46         if value == countMax:
47             print("测试数据为:",test,"   分类结果为：",key)
48     
49 g,l=creatDataSet( )
50 Classify_By_KNN( g , l )