Python3 k-邻近算法（KNN）

 # -*- coding: utf-8 -*-

 """

 Created on Fri Dec 29 13:13:44 2017

 @author: markli

 """

 import numpy as np;

 #两点之间的距离采用欧式几何距离

 '''

 采用欧式距离进行K最小临近分类

 x 未知分类点 m*1 向量

 y n个测试样本点 m*n 维向量

 '''

 def ComputeDistance(x,y):

     m = len(x); #获取维度数量

     #print(m);

     tempeye = -np.eye(m);

     tempone = np.ones((1,m));

     C = np.vstack((tempone,tempeye));#中间过渡矩阵 m+1 * m 按列合并，列数不变扩张行

     translateMatrix = np.hstack((x,y)); #按行合并，行数不变，扩张列

     tempresult = np.dot(translateMatrix,C);

     result = np.multiply(tempresult,tempresult);

     #result = [d**2 for d in np.array(tempresult)];

     result = np.sum(result,axis=0)

     distance = [pow(d,1/m) for d in np.array(result)];

     return distance;

 '''

 k 选取点的个数

 distance 带预测点与每个样本点的距离

 labels 每个样本点的类别标记

 return 返回距离最近的k的样本点的类别标记

 '''

 def KNN(k,distance,labels):

     dis_label = [];

     for i in range(len(labels)):

         tup = (distance[i],labels[i]);

         dis_label.append(tup);

     dis_label = sorted(dis_label,lambda x:x[0]);

     Kmin = [];

     for i in range(k-1):

         label = dis_label[i][1];

         if label not in Kmin:

             Kmin.append(label);

     return Kmin;

 #sklearn 中的KNN

 # -*- coding: utf-8 -*-

 """

 Created on Sat Dec 30 09:36:18 2017

 @author: markli

 """

 from sklearn import neighbors;

 from sklearn import datasets;

 import numpy as np;

 import matplotlib.pyplot as plt;

 KNN = neighbors.KNeighborsClassifier(n_neighbors=5,weights='distance');

 iris = datasets.load_iris();

 #print(iris);

 KNN.fit(iris.data,iris.target);

 x = [0.2,0.4,0.3,0.5];

 y = KNN.predict(np.array(x).reshape((1,4)));

 print(iris.target_names[y]);

 #k = neighbors.NearestNeighbors();

 #A = k.kneighbors_graph(iris.data,n_neighbors=5,mode='distance');