python实现Apriori**
本帖最后由 糖逗 于 2020-11-18 15:35 编辑参考书籍：《机器学习实战》
import numpy as np
def loadDataSet():
return [, , , ]
def createC1(dataSet):
C1 = []
for transaction in dataSet:
for item in transaction:#遍历list中的每个list中的元素
if not in C1:#为了使用frozenset对每个元素映射，用
C1.append()
C1.sort()#从小到大排序
#https://www.runoob.com/python/python-func-frozenset.html
return list(map(frozenset, C1))#之后要将集合作为字典键使用，frozenset可以实现而set不行
def scanD(D, Ck, minSupport):
ssCnt = {}
for tid in D:
for can in Ck:
#https://www.runoob.com/python3/ref-set-issubset.html
if can.issubset(tid):
if can not in ssCnt:
ssCnt = 1
else:
ssCnt += 1
numItems = float(len(D))
retList = []
supportData = {}
for key in ssCnt:#遍历每个键
support = ssCnt/numItems
if support >= minSupport:
retList.insert(0, key)#每次在最前面加入
supportData = support
return retList, supportData#返回满足支持度阈值的元素和所有元素支持度
def aprioriGen(Lk, k):
retList = []
lenLk = len(Lk)
for i in range(lenLk):
for j in range(i + 1, lenLk):
L1 = list(Lk)[: k - 2]#前k-2个相同时，将两个集合合并，这样做会漏项吧？？
L2 = list(Lk)[: k - 2]
L1.sort()
L2.sort()
if L1 == L2:
retList.append(Lk | Lk)#集合合并
return retList
def apriori(dataSet, minSupport = 0.5):
C1 = createC1(dataSet)
D = list(map(set, dataSet))
L1, supportData = scanD(D, C1, minSupport)
L =
k = 2#作为L中的index，k生成的组合数
while (len(L) > 0):
Ck = aprioriGen(L, k)
Lk, supK = scanD(D, Ck, minSupport)
supportData.update(supK)
L.append(Lk)
k += 1
return L, supportData
def generateRules(L, supportData, minConf = 0.7):#生成关联规则
bigRuleList = []#L已经经过了第一次minsupport筛选了，相当于剪枝
for i in range(1, len(L)):
for freqSet in L:
H1 = ) for item in freqSet]
if (i > 1):#元素超过两个以上的需要建立规则
rulesFromConseq(freqSet, H1, supportData, bigRuleList, minConf)
else:
calcConf(freqSet, H1, supportData, bigRuleList, minConf)
return bigRuleList
def calcConf(freqSet, H, supportData, brl, minConf = 0.7):#可信度值计算，H是freqSet的子集
prunedH = []
for conseq in H:
conf = supportData/supportData
if conf >= minConf:
print(freqSet-conseq,'-->',conseq,'conf:', conf)
brl.append((freqSet-conseq, conseq, conf))
prunedH.append(conseq)
return prunedH
def rulesFromConseq(freqSet, H, supportData, brl, minConf = 0.7):
m = len(H)
if (len(freqSet) > (m + 1)): #因为后文需要将集合合并成一个每组m+1的集合，所以m+1要小于freqSet
Hmp1 = aprioriGen(H, m + 1)
Hmp1 = calcConf(freqSet, Hmp1, supportData, brl, minConf)
if (len(Hmp1) > 1):#假设2->1,3 ; 3->1,2;则想进一步验证2,3 -> 1
rulesFromConseq(freqSet, Hmp1, supportData, brl, minConf)
return
if __name__ == "__main__":
dataSet = loadDataSet()
L, suppData = apriori(dataSet, minSupport = 0.5)
rules = generateRules(L, suppData, minConf = 0.7) 需要再看一遍，这个代码有点复杂。{:10_255:} 宝藏鱼友啊{:10_297:} 昨非 发表于 2020-11-18 15:35
宝藏鱼友啊
好书和好内容应该大家一起分享嘛{:10_259:} 糖逗 发表于 2020-11-18 15:37
好书和好内容应该大家一起分享嘛
鉴定完毕，盲猜是竞赛大佬
膜！！{:10_301:} 昨非 发表于 2020-11-18 15:39
鉴定完毕，盲猜是竞赛大佬
膜！！
不是的，这是书里代码{:10_255:} 糖逗 发表于 2020-11-18 19:29
不是的，这是书里代码
{:10_245:}
页:
[1]