python調用R語言，關聯規則可視化

首先當然要配置r語言環境變量什么的

D:\R-3.5.1\bin\x64;
D:\R-3.5.1\bin\x64\R.dll;
D:\R-3.5.1;
D:\ProgramData\Anaconda3\Lib\site-packages\rpy2;

 

本來用python也可以實現關聯規則，雖然沒包，但是可視化挺麻煩的

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
from pandas import read_csv



def loadDataSet():
    dataset = read_csv("F:/goverment/Aprior/No Number.csv")
    data = dataset.values[:,:]
    Data=[]
    for line in data:
        ls=[]
        for i in line:
            ls.append(i)
        Data.append(ls)
    #print(Data)
    return Data

    '''
    return [['a', 'c', 'e'], ['b', 'd'], ['b', 'c'], ['a', 'b', 'c', 'd'], ['a', 'b'], ['b', 'c'], ['a', 'b'],
            ['a', 'b', 'c', 'e'], ['a', 'b', 'c'], ['a', 'c', 'e']]'''

def createC1(dataSet):
    C1 = []
    for transaction in dataSet:
        for item in transaction:
            if not [item] in C1:
                C1.append([item])
    C1.sort()
    '''??????????????????????????????????????????????????????'''
    # 映射為frozenset唯一性的，可使用其構造字典
    return list(map(frozenset, C1))      
 
 
# 從候選K項集到頻繁K項集（支持度計算）
def scanD(D, Ck, minSupport):
    ssCnt = {}
    for tid in D:
        for can in Ck:
            if can.issubset(tid):
                if not can in ssCnt:
                    ssCnt[can] = 1
                else:
                    ssCnt[can] += 1
    numItems = float(len(D))
    retList = []
    supportData = {}
    for key in ssCnt:
        support = ssCnt[key] / numItems
        if support >= minSupport:
            retList.insert(0, key)
            supportData[key] = support  
    return retList, supportData
 
 
def calSupport(D, Ck, min_support):
    dict_sup = {}
    for i in D:
        for j in Ck:
            if j.issubset(i):
                if not j in dict_sup:
                    dict_sup[j] = 1
                else:
                    dict_sup[j] += 1
    sumCount = float(len(D))
    supportData = {}
    relist = []
    for i in dict_sup:
        temp_sup = dict_sup[i] / sumCount
        if temp_sup >= min_support:
            relist.append(i)
            supportData[i] = temp_sup  # 此處可設置返回全部的支持度數據（或者頻繁項集的支持度數據）
    return relist, supportData
 
 
# 改進剪枝算法
def aprioriGen(Lk, k):  # 創建候選K項集 ##LK為頻繁K項集
    retList = []
    lenLk = len(Lk)
    for i in range(lenLk):
        for j in range(i + 1, lenLk):
            L1 = list(Lk[i])[:k - 2]
            L2 = list(Lk[j])[:k - 2]
            L1.sort()
            L2.sort()
            if L1 == L2:  # 前k-1項相等，則可相乘，這樣可防止重復項出現
                #  進行剪枝（a1為k項集中的一個元素，b為它的所有k-1項子集）
                a = Lk[i] | Lk[j]  # a為frozenset()集合
                a1 = list(a)
                b = []
                # 遍歷取出每一個元素，轉換為set，依次從a1中剔除該元素，並加入到b中
                for q in range(len(a1)):
                    t = [a1[q]]
                    tt = frozenset(set(a1) - set(t))
                    b.append(tt)
                t = 0
                for w in b:
                    # 當b（即所有k-1項子集）都是Lk（頻繁的）的子集，則保留，否則刪除。
                    if w in Lk:
                        t += 1
                if t == len(b):
                    retList.append(b[0] | b[1])
    return retList
 
 
def apriori(dataSet, minSupport=0.2):
    C1 = createC1(dataSet)
    D = list(map(set, dataSet))  # 使用list()轉換為列表
    L1, supportData = calSupport(D, C1, minSupport)
    L = [L1]  # 加列表框，使得1項集為一個單獨元素
    k = 2
    while (len(L[k - 2]) > 0):
        Ck = aprioriGen(L[k - 2], k)
        Lk, supK = scanD(D, Ck, minSupport)  # scan DB to get Lk
        supportData.update(supK)
        L.append(Lk)  # L最后一個值為空集
        k += 1
    del L[-1]  # 刪除最后一個空集
    return L, supportData  # L為頻繁項集，為一個列表，1，2，3項集分別為一個元素。
 
 
# 生成集合的所有子集
def getSubset(fromList, toList):
    for i in range(len(fromList)):
        t = [fromList[i]]
        tt = frozenset(set(fromList) - set(t))
        if not tt in toList:
            toList.append(tt)
            tt = list(tt)
            if len(tt) > 1:
                getSubset(tt, toList)
 
 
#def calcConf(freqSet, H, supportData, ruleList, minConf=0.7):
def calcConf(freqSet, H, supportData, Rule, minConf=0.7):
    for conseq in H:
        conf = supportData[freqSet] / supportData[freqSet - conseq]  # 計算置信度
        # 提升度lift計算lift = p(a & b) / p(a)*p(b)
        lift = supportData[freqSet] / (supportData[conseq] * supportData[freqSet - conseq])
        
        ls=[]
        if conf >= minConf and lift > 3:
            for i in freqSet - conseq:
                #print(i," ",end="")
                ls.append(i)
                ls.append(" ")
            #print('-->',end="")
            ls.append('-->')
            for i in conseq:
                #print(i," ",end="")
                ls.append(i)
                ls.append(" ")
            #print('支持度:', round(supportData[freqSet - conseq]*100, 1), "%",'  置信度:', round(conf*100,1),"%",'  lift值為', round(lift, 2))
            #ls.append(' 支持度:')
            #ls.append(round(supportData[freqSet - conseq]*100, 1))
            #ls.append("% ")
            #ls.append(' 置信度:')
            ls.append( round(conf*100,1))
            ls.append("% ")
            #ls.append( round(lift, 2))
            #ls.append(round(lift, 2))
            
            #ruleList.append((freqSet - conseq, conseq, conf))
        if ls!=[]: 
            #print(len(ls))
            Rule.append(ls)
# =============================================================================
#     for line in Rule:
#         for i in line:
#             print(i,end="")
#         print("")
# =============================================================================
    return Rule
# =============================================================================
#             print(freqSet - conseq, '-->', conseq, '支持度', round(supportData[freqSet - conseq], 2), '置信度：', round(conf,3),
#                   'lift值為：', round(lift, 2))
# =============================================================================
            
 
# 生成規則
def gen_rule(L, supportData, minConf=0.7):
    bigRuleList = []
    for i in range(1, len(L)):  # 從二項集開始計算
        for freqSet in L[i]:  # freqSet為所有的k項集
            # 求該三項集的所有非空子集，1項集，2項集，直到k-1項集，用H1表示，為list類型,里面為frozenset類型，
            H1 = list(freqSet)
            all_subset = []
            getSubset(H1, all_subset)  # 生成所有的子集
            calcConf(freqSet, all_subset, supportData, bigRuleList, minConf)
    return bigRuleList
 
 
if __name__ == '__main__':
    
    dataSet = loadDataSet()
    #print(dataSet)
    L, supportData = apriori(dataSet, minSupport=0.05)
    rule = gen_rule(L, supportData, minConf=0.5)
    for i in rule:
        for j in i:
            if j==',':
                continue
            else:
                print(j,end="")
        print("")

'''
具體公式：

P(B|A)/P(B)

稱為A條件對於B事件的提升度,如果該值=1,說明兩個條件沒有任何關聯,
如果<1,說明A條件(或者說A事件的發生)與B事件是相斥的,
一般在數據挖掘中當提升度大於3時,我們才承認挖掘出的關聯規則是有價值的。
'''

之后還是用r吧，要下載rpy2，見https://www.cnblogs.com/caiyishuai/p/9520214.html 

還要下載兩個R的包

import rpy2.robjects as robjects

b=('''
    install.packages("arules")
    install.packages("arulesViz")
''')
robjects.r(b)

然后就是主代碼了

import rpy2.robjects as robjects

a=('''Encoding("UTF-8")
setwd("F:/goverment/Aprior")

all_data<-read.csv("F:/goverment/Aprior/NewData.csv",header = T,#將數據轉化為因子型
                   colClasses=c("factor","factor","factor","factor","factor","factor","factor","factor","factor","factor","factor","factor"))
library(arules)
rule=apriori(data=all_data[,c(1,4,5,6,7,8,9,10,12)], parameter = list(support=0.05,confidence=0.7,minlen=2,maxlen=10))       
   ''')
robjects.r(a)

robjects.r('''                        
rule.subset<-subset(rule,lift>1)
#inspect(rule.subset)
rules.sorted<-sort(rule.subset,by="lift")
subset.matrix<-is.subset(rules.sorted,rules.sorted)
lower.tri(subset.matrix,diag=T)
subset.matrix[lower.tri(subset.matrix,diag = T)]<-NA
redundant<-colSums(subset.matrix,na.rm = T)>=1          #這五條就是去冗余（感興趣可以去網上搜），我雖然這里寫了，但我沒有去冗余，我的去了以后一個規則都沒了
which(redundant)
rules.pruned<-rules.sorted[!redundant]
#inspect(rules.pruned) #輸出去冗余后的規則
''')


c=('''

library(arulesViz)#掉包

jpeg(file="plot1.jpg")
#inspect(rule.subset)
plt<-plot(rule.subset,shading = "lift")#畫散點圖
dev.off()


subrules<-head(sort(rule.subset,by="lift"),50)
#jpeg(file="plot2.jpg")
plot(subrules,method = "graph")#畫圖
#dev.off()

rule.sorted <- sort(rule.subset, decreasing=TRUE,  by="lift") #按提升度排序
rules.write<-as(rule.sorted,"data.frame") #將規則轉化為data類型
write.csv(rules.write,"F:/goverment/Aprior/NewRules.csv",fileEncoding="UTF-8")
''')
robjects.r(c)

#取出保存的規則，放到一個列表中
from pandas import read_csv
data_set = read_csv("F:/goverment/Aprior/NewRules.csv")
data = data_set.values[:, :]
rul = []
for line in data:
    ls = []
    for j in line:
        try :
            j=float(j)
            if j>0 and j<=1:
                j=str(round(j*100,2))+"%"
                ls.append(j)
            else:
                ls.append(round(j,2))
        except:
            ls.append(j)
    rul.append(ls)


for line in rul:
    print(line)