本博客介绍了一种使用卡方检验评估中药性味归经与功能之间关系的方法，并通过Python脚本实现了对多元素组合的数据挖掘过程。
sklearn实战-乳腺癌细胞数据挖掘(博客主亲自录制视频教程)
https://study.163.com/course/introduction.htm?courseId=1005269003&utm_campaign=commission&utm_source=cp-400000000398149&utm_medium=share
脚本名称
censor_relation_xwgj2_functions.py
需要导入中药表excel
# -*- coding: utf-8 -*-
"""
Created on Thu Aug 25 10:06:52 2016
审核性味归经与功能的脚本(超过四元素)
@author: Administrator
"""
import xlrd,csv
import chi_square,list_xwgj_function_multiElements
#多元素组合
thelist_xwgj_function=list_xwgj_function_multiElements.list_xwgj2_function
#数据处理，把['寒,苦,肝,胃', '清热']转换为[['寒','苦','肝','胃'], '清热']
thelist_xwgj_function=[[i[0].split(","),i[1]] for i in thelist_xwgj_function]
fileName="性味归经_功能_关系_多元素.csv"
excelFilename="中药表.xlsx"
sheetName="Sheet1"
#打开excel数据
excelFile=xlrd.open_workbook(excelFilename)
sheet=excelFile.sheet_by_name(sheetName)
#表单行与列
number_rows=sheet.nrows
number_columns=sheet.ncols
#表单内数据
list_sheetData=[]
#
list_xingWeiFunction_relation=[]
#获取表格数据
def Get_sheetData():
for row in range(1,number_rows):
list_sheetData.append(sheet.row_values(row))
return list_sheetData
list_sheetData=Get_sheetData()
#,卡方统计的a是否成立，实验组True，对照组True
#例如keyWord1=['寒', '苦', '肝', '胃']，keyWord2='清热'
#i为['苦，寒归胃经\r\n\t', '清热解毒，消痈，下乳，舒筋通脉。']
def True_a(keyWord1,keyWord2,i):
#key为['寒', '苦', '肝', '胃']内元素
for key in keyWord1:
if key not in i[0]:
return False
if keyWord2 not in i[1]:
return False
return True
#,卡方统计的b是否成立,实验组True，对照组False
#例如keyWord1=['寒', '苦', '肝', '胃']，keyWord2='清热'
#i为['苦，寒归胃经\r\n\t', '清热解毒，消痈，下乳，舒筋通脉。']
def True_b(keyWord1,keyWord2,i):
#key为['寒', '苦', '肝', '胃']内元素
for key in keyWord1:
if key not in i[0]:
return False
if keyWord2 in i[1]:
return False
return True
#,卡方统计的c是否成立，实验组False，对照组True
#例如keyWord1=['寒', '苦', '肝', '胃']，keyWord2='清热'
#i为['苦，寒归胃经\r\n\t', '清热解毒，消痈，下乳，舒筋通脉。']
def True_c(keyWord1,keyWord2,i):
#只要count不等于len(keyWord1),就表示keyWord1至少有一个元素不符合
len_keyWord1=len(keyWord1)
count=0
#key为['寒', '苦', '肝', '胃']内元素
for key in keyWord1:
if key in i[0]:
count+=1
if len_keyWord1==count:
return False
if keyWord2 not in i[1]:
return False
return True
#,卡方统计的d是否成立，实验组False，对照组False
#例如keyWord1=['寒', '苦', '肝', '胃']，keyWord2='清热'
#i为['苦，寒归胃经\r\n\t', '清热解毒，消痈，下乳，舒筋通脉。']
def True_d(keyWord1,keyWord2,i):
#只要count不等于len(keyWord1),就表示keyWord1至少有一个元素不符合
len_keyWord1=len(keyWord1)
count=0
#key为['寒', '苦', '肝', '胃']内元素
for key in keyWord1:
if key in i[0]:
count+=1
if len_keyWord1==count:
return False
if keyWord2 in i[1]:
return False
return True
#卡方统计的keyWord1与keyWord2的关系
#三元素和多元素的脚本不一样，此函数需要修改
def Get_single_chiSquare(keyWord1,keyWord2):
#记录包含卡方值a的uid号，方便逆向检测
list_a=[]
a=0
b=0
c=0
d=0
#统计包含a数量
for k in range(len(list_sheetData)):
i=list_sheetData[k]
if True_a(keyWord1,keyWord2,i):
a+=1
list_a.append(k+2)
#统计包含b数量
for i in list_sheetData:
if True_b(keyWord1,keyWord2,i):
b+=1
#统计包含c数量
for i in list_sheetData:
if True_c(keyWord1,keyWord2,i):
c+=1
#统计包含d数量
for i in list_sheetData:
if True_d(keyWord1,keyWord2,i):
d+=1
#print("a:",a)
#print("b:",b)
#print("c:",c)
#print("d:",d)
k2=chi_square.value_independence(a,b,c,d)
relation=chi_square.judge_independence(a,b,c,d)
lis1=[[a,b,c,d],k2,relation,list_a]
return lis1
#计算所有性味的相关性
def Get_all_chiSquare(thelist_xwgj_function):
for i in thelist_xwgj_function:
try:
keyWord1,keyWord2=i[0],i[1]
relation=Get_single_chiSquare(keyWord1,keyWord2)
list1=[i,relation,relation[2]]
list_xingWeiFunction_relation.append(list1)
except:
print("wrong at:",i)
continue
return list_xingWeiFunction_relation
def Write_table_to_csv(list1,fileName):
#对列表格式修改，字符串写入的格式不对
file=open(fileName,'w',newline='')
writer1=csv.writer(file)
writer1.writerows(list1)
file.close()
list_xingWeiFunction_relation=Get_all_chiSquare(thelist_xwgj_function)
Write_table_to_csv(list_xingWeiFunction_relation,fileName)
'''
#测试数据
i=['苦，寒,胃经\r\n\t', '清热，解毒，消痈，下乳，舒筋通脉。']
b=thelist_xwgj_function[0]
keyWord1,keyWord2=b[0],b[1]
'''
chi_square脚本
#coding=utf-8
#独立性检验test for independence,也是卡方检验chi_square
#前提条件：a,b,c,d 必须大于5
#2.706是判断标准（90概率），值越大，越有关，值越小，越无关
def value_independence(a,b,c,d):
if a>=5 and b>=5 and c>=5 and d>=5:
return ((a+b+c+d)*(a*d-b*c)**2)/float((a+b)*(c+d)*(a+c)*(b+d))
#返回True表示有关
#返回False表示无关
#2.706表示有90%概率
def judge_independence(a,b,c,d):
num_independence=value_independence(a,b,c,d)
print("chi_square:",num_independence)
if num_independence==None:
print("not suit for chi_square statistics")
return "wrong"
if num_independence>2.706:
print ("there is relationship")
return True
if num_independence<=2.706:
print("there is no relationship")
return False
list_xwgj_function_multiElements 脚本# -*- coding: utf-8 -*-
"""
Created on Tue Aug 23 11:38:35 2016
生成多个因素的性味归经组合，四位，五位
算法：卡方
@author: Administrator
"""
import xlrd,csv
#性味归经的元数据
list_xing=["寒","凉","温","热"]
list_wei=["酸","苦","甘","辛","咸"]
list_guiJing=["胃","肝","胆","心","肺","肾","脾","膀胱","大肠","小肠"]
#四元素以上的性味归经功能吧
excelFilename="性味归经清理3.xlsx"
excelFilename1="性味归经清理4.xlsx"
sheetName="Sheet1"
#打开excel数据
excelFile=xlrd.open_workbook(excelFilename)
sheet=excelFile.sheet_by_name(sheetName)
#表单行与列
number_rows=sheet.nrows
number_columns=sheet.ncols
#表格性味归经
list_xwg2j=sheet.col_values(0)[1:]
#功能
list_function=sheet.col_values(1)[1:]
list_function1=[i for i in list_function if i!=""]
#四元素以上性味归经与功能组合
list_xwgj2_function=[]
#生成两个归经的组合
def Get_list_xwg2j():
for i in list_xwg2j:
for k in list_guiJing:
if k not in i:
element=i+","+k
list_xwg2j.append(element)
#print ("i:",i)
return list_xwg2j
#添加一些元素
def add_some_elements(list_xwg2j):
list_xwg2j.append('甘,温,肝,肾')
list_xwg2j.append('甘,温,脾,胃')
list_xwg2j.append('苦,寒,肝,胃,大肠')
list_xwg2j.append('辛,温,脾,胃,肺')
list_xwg2j.append('苦,寒,大肠,肺')
list_xwg2j.append('苦,寒,心,肝')
return list_xwg2j
#写入csv前准备，添加序列号
def add_uid(list_xwgj_clean):
list1=[]
for i in range(len(list_xwgj_clean)):
value=list_xwgj_clean[i]
item=[i,value]
list1.append(item)
return list1
def Write_table_to_csv(list1,fileName):
#对列表格式修改，字符串写入的格式不对
file=open(fileName,'w',newline='')
writer1=csv.writer(file)
writer1.writerows(list1)
file.close()
#性味归经大于四的多元素组合
def Get_list_xwgj2_function(list1):
for i in list1:
for k in list_function1:
list_xwgj2_function.append([i,k])
return list_xwgj2_function
#list_xwg2j=Get_list_xwg2j()
#list_xwg2j_add=add_some_elements(list_xwg2j)
#写入csv前准备,添加序列号
#list_write=add_uid(list_xwg2j_add)
#Write_table_to_csv(list_write,"性味归经清理3.csv")
list_xwgj2_function=Get_list_xwgj2_function(list_xwg2j)
list_xwgj_function_multiElements 脚本 需要性味归经清理3excel
结果
计算出细化的性味归经组合，从三万多组合中，得到472种组合
扫一扫
2157
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
