灰色关联度算法

转自https://blog.csdn.net/fontthrone/article/details/80607794 Font Tian同学

点击(此处)折叠或打开

1. #coding:utf-8
   
2. import sys
   
3. import pandas as pd
4. import numpy as np
   
5. from numpy import *
   
6. import matplotlib.pyplot as plt
   
7. 
   
8. 
   
9. 
   
10. def GRA_ONE(gray, m=0):
    
11.     # 读取为df格式
    
12.     gray = (gray - gray.min()) / (gray.max() - gray.min())
    
13.     # 标准化
    
14.     std = gray.iloc[:, m] # 为标准要素
    
15.     ce = gray.iloc[:, 0:] # 为比较要素
    
16.     n, m = ce.shape[0], ce.shape[1] # 计算行列
    
17.     # 与标准要素比较，相减
    
18.     a = zeros([m, n])
    
19.     for i in range(m):
    
20.         for j in range(n):
    
21.             a[i, j] = abs(ce.iloc[j, i] - std[j])
    
22.     # 取出矩阵中最大值与最小值
    
23.     c, d = amax(a), amin(a)
    
24.     # 计算值
    
25.     result = zeros([m, n])
    
26.     for i in range(m):
    
27.         for j in range(n):
    
28.             ss = a[i, j] + 0.5 * c
    
29.             if ss ==0:
    
30.                 result[i, j] =1
    
31.             else:
    
32.                 result[i, j] = (d + 0.5 * c) / (a[i, j] + 0.5 * c)
    
33. 
    
34.     # 求均值，得到灰色关联值,并返回
    
35.     return pd.DataFrame([mean(result[i, :]) for i in range(m)])
    
36. 
    
37. 
    
38. def GRA(DataFrame):
    
39.     list_columns = [
    
40.         str(s) for s in range(len(DataFrame.columns)) if s not in [None]
    
41.     ]
    
42.     df_local = pd.DataFrame(columns=list_columns)
    
43.     for i in range(len(DataFrame.columns)):
    
44.         df_local.iloc[:, i] = GRA_ONE(DataFrame, m=i)[0]
    
45.     return df_local
    
46. 
    
47. wine = pd.read_csv("h2.csv")
    
48. wine.head()
    
49. print "----------------------------"
    
50. data_wine_gra = GRA(wine)
    
51. 
    
52. print data_wine_gra

附件
h.csv

点击(此处)折叠或打开

1. a,b,c ,d
   
2. 1,2,1,1
   
3. 2,2,2,1
   
4. 3,3,3,2
   
5. 4,2,4,2
   
6. 5,4,5,3
   
7. 6,1,6,3

h2.csv

点击(此处)折叠或打开

1. a,b
   
2. 1,100
   
3. 2,200
   
4. 3,300
   
5. 4,400
   
6. 5,500
   
7. 6,491

该输出为矩阵 [i,j] 表示第i列和第j列关联度