NMI計算

 

 

NMI計算

NMI(Normalized Mutual Information)標准化互信息，常用在聚類中，度量兩個聚類結果的相近程度。是社區發現(community detection)的重要衡量指標，基本可以比較客觀地評價出一個社區划分與標准划分之間相比的准確度。NMI的值域是0到1，越高代表划分得越准。

# -*- coding:utf-8 -*-
'''
Created on 2017年10月28日

@summary: 利用Python實現NMI計算

@author: dreamhome
'''
import math
import numpy as np
from sklearn import metrics
def NMI(A,B):
    #樣本點數
    total = len(A)
    A_ids = set(A)
    B_ids = set(B)
    #互信息計算
    MI = 0
    eps = 1.4e-45
    for idA in A_ids:
        for idB in B_ids:
            idAOccur = np.where(A==idA)
            idBOccur = np.where(B==idB)
            idABOccur = np.intersect1d(idAOccur,idBOccur)
            px = 1.0*len(idAOccur[0])/total
            py = 1.0*len(idBOccur[0])/total
            pxy = 1.0*len(idABOccur)/total
            MI = MI + pxy*math.log(pxy/(px*py)+eps,2)
    # 標准化互信息
    Hx = 0
    for idA in A_ids:
        idAOccurCount = 1.0*len(np.where(A==idA)[0])
        Hx = Hx - (idAOccurCount/total)*math.log(idAOccurCount/total+eps,2)
    Hy = 0
    for idB in B_ids:
        idBOccurCount = 1.0*len(np.where(B==idB)[0])
        Hy = Hy - (idBOccurCount/total)*math.log(idBOccurCount/total+eps,2)
    MIhat = 2.0*MI/(Hx+Hy)
    return MIhat

if __name__ == '__main__':
    A = np.array([1,1,1,1,1,1,2,2,2,2,2,2,3,3,3,3,3])
    B = np.array([1,2,1,1,1,1,1,2,2,2,2,3,1,1,3,3,3])
    print NMI(A,B)
    print metrics.normalized_mutual_info_score(A,B)


原文：https://blog.csdn.net/DreamHome_S/article/details/78379635 

 

# coding=utf-8
import numpy as np
import math
def NMI(A,B):
    # len(A) should be equal to len(B)
    total = len(A)
    A_ids = set(A)
    B_ids = set(B)
    #Mutual information
    MI = 0
    eps = 1.4e-45
    for idA in A_ids:
        for idB in B_ids:
            idAOccur = np.where(A==idA)
            idBOccur = np.where(B==idB)
            idABOccur = np.intersect1d(idAOccur,idBOccur)
            px = 1.0*len(idAOccur[0])/total
            py = 1.0*len(idBOccur[0])/total
            pxy = 1.0*len(idABOccur)/total
            MI = MI + pxy*math.log(pxy/(px*py)+eps,2)
    # Normalized Mutual information
    Hx = 0
    for idA in A_ids:
        idAOccurCount = 1.0*len(np.where(A==idA)[0])
        Hx = Hx - (idAOccurCount/total)*math.log(idAOccurCount/total+eps,2)
    Hy = 0
    for idB in B_ids:
        idBOccurCount = 1.0*len(np.where(B==idB)[0])
        Hy = Hy - (idBOccurCount/total)*math.log(idBOccurCount/total+eps,2)
    MIhat = 2.0*MI/(Hx+Hy)
    return MIhat

if __name__ == '__main__':
    A = np.array([1,1,1,1,1,1,2,2,2,2,2,2,3,3,3,3,3])
    B = np.array([1,2,1,1,1,1,1,2,2,2,2,3,1,1,3,3,3])
    print (NMI(A,B))

結果：0.36456

這一篇博文寫的不錯

 

 

 

自己編寫了一個，同時做了排序處理

# coding=utf-8
import numpy as np
import math
import operator


def NMI(A,B):
    # len(A) should be equal to len(B)
    total = len(A)
    A_ids = set(A)
    B_ids = set(B)
    #Mutual information
    MI = 0
    eps = 1.4e-45
    for idA in A_ids:
        for idB in B_ids:
            idAOccur = np.where(A==idA)
            idBOccur = np.where(B==idB)
            idABOccur = np.intersect1d(idAOccur,idBOccur)
            px = 1.0*len(idAOccur[0])/total
            py = 1.0*len(idBOccur[0])/total
            pxy = 1.0*len(idABOccur)/total
            MI = MI + pxy*math.log(pxy/(px*py)+eps,2)
    # Normalized Mutual information
    Hx = 0
    for idA in A_ids:
        idAOccurCount = 1.0*len(np.where(A==idA)[0])
        Hx = Hx - (idAOccurCount/total)*math.log(idAOccurCount/total+eps,2)
    Hy = 0
    for idB in B_ids:
        idBOccurCount = 1.0*len(np.where(B==idB)[0])
        Hy = Hy - (idBOccurCount/total)*math.log(idBOccurCount/total+eps,2)
    MIhat = 2.0*MI/(Hx+Hy)
    return MIhat


if __name__ == '__main__':
    A = np.array([1,1,1])
    B = np.array([2,3,4])
    C = np.array([1,1,6])
    print(NMI(A,B))
    m=[]#包含了位置的互信息
    n=[]#只有互信息
    dic={}
    q=1
    m.append(NMI(A,B))
    m.append(NMI(B,C))
    m.append(NMI(A,C))
    
    
    for i in m:
        dic['第{}個互信息'.format(q)]='{}'.format(i)
        q=q+1
    print(dic)
    rankdata=sorted(dic.items(),key=operator.itemgetter(1),reverse=True)
    print(rankdata)
    
    

實驗結果如圖