1 # -*- encoding=utf-8 -*-
 2
 3 import warnings
 4 warnings.filterwarnings("ignore")
 5 import pandas as pd
 6 import numpy as np
 7
 8
 9 def get_score(wi_list,data):
10     """
11     :param wi_list: 权重系数列表
12     :param data：评价指标数据框
13     :return:返回得分
14     """
15
16     #  将权重转换为矩阵
17
18     cof_var = np.mat(wi_list)
19
20     #  将数据框转换为矩阵
21     context_train_data = np.mat(data)
22
23     #  权重跟自变量相乘
24     last_hot_matrix = context_train_data * cof_var.T
25     last_hot_matrix = pd.DataFrame(last_hot_matrix)
26
27     #  累加求和得到总分
28     last_hot_score = list(last_hot_matrix.apply(sum))
29
30     #  max-min 归一化
31
32     # last_hot_score_autoNorm = autoNorm(last_hot_score)
33
34     # 值映射成分数（0-100分）
35
36     # last_hot_score_result = [i * 100 for i in last_hot_score_autoNorm]
37
38     return last_hot_score
39
40
41
42 def get_entropy_weight(data):
43     """
44     :param data: 评价指标数据框
45     :return: 各指标权重列表
46     """
47     # 数据标准化
48     data = (data - data.min())/(data.max() - data.min())
49     m,n=data.shape
50     #将dataframe格式转化为matrix格式
51     data=data.as_matrix(columns=None)
52     k=1/np.log(m)
53     yij=data.sum(axis=0)
54     #第二步，计算pij
55     pij=data/yij
56     test=pij*np.log(pij)
57     test=np.nan_to_num(test)
58
59     #计算每种指标的信息熵
60     ej=-k*(test.sum(axis=0))
61     #计算每种指标的权重
62     wi=(1-ej)/np.sum(1-ej)
63
64     wi_list=list(wi)
65
66
67     return  wi_list
68
69
70
71 if __name__ == '__main__':
72
73
74     data0 = pd.read_excel("C:\\Users\\Oreo\\Desktop\\test2.xlsx", encoding='utf8')
75
76     data = data0.iloc[:, 1:10]
77     mm=data
78     wi_list=get_entropy_weight(data)
79     score_list=get_score(mm,wi_list)
80     mm['score']=score_list
81     mm['科室']=data0['科室']
82     # 然后对数据框按得分从大到小排序
83     result = mm.sort_values(by='score', axis=0, ascending=False)
84     result['rank'] = range(1, len(result) + 1)
85
86     print(result)
87
88     # 写出csv数据
89     result.to_csv('C:\\Users\\Oreo\\Desktop\\test2_result.csv', index=False)