python3randomlinux熵_Python计算信息熵实例

计算信息熵的公式&＃xff1a;n是类别数&＃xff0c;p(xi)是第i类的概率

假设数据集有m行&＃xff0c;即m个样本&＃xff0c;每一行最后一列为该样本的标签&＃xff0c;计算数据集信息熵的代码如下&＃xff1a;

from math import log

def calcShannonEnt(dataSet):

numEntries &＃61; len(dataSet) # 样本数

labelCounts &＃61; {} # 该数据集每个类别的频数

for featVec in dataSet: # 对每一行样本

currentLabel &＃61; featVec[-1] # 该样本的标签

if currentLabel not in labelCounts.keys(): labelCounts[currentLabel] &＃61; 0

labelCounts[currentLabel] &＃43;&＃61; 1

shannonEnt &＃61; 0.0

for key in labelCounts:

prob &＃61; float(labelCounts[key])/numEntries # 计算p(xi)

shannonEnt -&＃61; prob * log(prob, 2) # log base 2

return shannonEnt

补充知识&＃xff1a;python 实现信息熵、条件熵、信息增益、基尼系数

我就废话不多说了&＃xff0c;大家还是直接看代码吧~

import pandas as pd

import numpy as np

import math

## 计算信息熵

def getEntropy(s):

# 找到各个不同取值出现的次数

if not isinstance(s, pd.core.series.Series):

s &＃61; pd.Series(s)

prt_ary &＃61; pd.groupby(s , by &＃61; s).count().values / float(len(s))

return -(np.log2(prt_ary) * prt_ary).sum()

## 计算条件熵: 条件s1下s2的条件熵

def getCondEntropy(s1 , s2):

d &＃61; dict()

for i in list(range(len(s1))):

d[s1[i]] &＃61; d.get(s1[i] , []) &＃43; [s2[i]]

return sum([getEntropy(d[k]) * len(d[k]) / float(len(s1)) for k in d])

## 计算信息增益

def getEntropyGain(s1, s2):

return getEntropy(s2) - getCondEntropy(s1, s2)

## 计算增益率

def getEntropyGainRadio(s1, s2):

return getEntropyGain(s1, s2) / getEntropy(s2)

## 衡量离散值的相关性

import math

def getDiscreteCorr(s1, s2):

return getEntropyGain(s1,s2) / math.sqrt(getEntropy(s1) * getEntropy(s2))

# ######## 计算概率平方和

def getProbSS(s):

if not isinstance(s, pd.core.series.Series):

s &＃61; pd.Series(s)

prt_ary &＃61; pd.groupby(s, by &＃61; s).count().values / float(len(s))

return sum(prt_ary ** 2)

######## 计算基尼系数

def getGini(s1, s2):

d &＃61; dict()

for i in list(range(len(s1))):

d[s1[i]] &＃61; d.get(s1[i] , []) &＃43; [s2[i]]

return 1-sum([getProbSS(d[k]) * len(d[k]) / float(len(s1)) for k in d])

## 对离散型变量计算相关系数&＃xff0c;并画出热力图, 返回相关性矩阵

def DiscreteCorr(C_data):

## 对离散型变量(C_data)进行相关系数的计算

C_data_column_names &＃61; C_data.columns.tolist()

## 存储C_data相关系数的矩阵

import numpy as np

dp_corr_mat &＃61; np.zeros([len(C_data_column_names) , len(C_data_column_names)])

for i in range(len(C_data_column_names)):

for j in range(len(C_data_column_names)):

# 计算两个属性之间的相关系数

temp_corr &＃61; getDiscreteCorr(C_data.iloc[:,i] , C_data.iloc[:,j])

dp_corr_mat[i][j] &＃61; temp_corr

# 画出相关系数图

fig &＃61; plt.figure()

fig.add_subplot(2,2,1)

sns.heatmap(dp_corr_mat ,vmin&＃61; - 1, vmax&＃61; 1, cmap&＃61; sns.color_palette(&＃39;RdBu&＃39; , n_colors&＃61; 128) , xticklabels&＃61; C_data_column_names , yticklabels&＃61; C_data_column_names)

return pd.DataFrame(dp_corr_mat)

if __name__ &＃61;&＃61; "__main__":

s1 &＃61; pd.Series([&＃39;X1&＃39; , &＃39;X1&＃39; , &＃39;X2&＃39; , &＃39;X2&＃39; , &＃39;X2&＃39; , &＃39;X2&＃39;])

s2 &＃61; pd.Series([&＃39;Y1&＃39; , &＃39;Y1&＃39; , &＃39;Y1&＃39; , &＃39;Y2&＃39; , &＃39;Y2&＃39; , &＃39;Y2&＃39;])

print(&＃39;CondEntropy:&＃39;,getCondEntropy(s1, s2))

print(&＃39;EntropyGain:&＃39; , getEntropyGain(s1, s2))

print(&＃39;EntropyGainRadio&＃39; , getEntropyGainRadio(s1 , s2))

print(&＃39;DiscreteCorr:&＃39; , getDiscreteCorr(s1, s1))

print(&＃39;Gini&＃39; , getGini(s1, s2))

以上这篇Python计算信息熵实例就是小编分享给大家的全部内容了&＃xff0c;希望能给大家一个参考&＃xff0c;也希望大家多多支持脚本之家。