1 def loadDataSet():
 2     return [[1,2,5],[2,4],[2,3],[1,2,4],[1,3],[2,3],[1,3],[1,2,3,5],[1,2,3]]
 3 #1.构建候选1项集C1
 4 def createC1(dataSet):
 5     C1 = []
 6     for transaction in dataSet:
 7         for item in transaction:
 8             if not [item] in C1:
 9                 C1.append([item])
10
11     C1.sort()
12     return list(map(frozenset, C1))
13
14 #将候选集Ck转换为频繁项集Lk
15 #D：原始数据集
16 #Cn: 候选集项Ck
17 #minSupport:支持度的最小值
18 def scanD(D, Ck, minSupport):
19     #候选集计数
20     ssCnt = {}
21     for tid in D:
22         for can in Ck:
23             if can.issubset(tid):
24                 if can not in ssCnt.keys(): ssCnt[can] = 1
25                 else: ssCnt[can] += 1
26
27     numItems = float(len(D))
28     Lk= []     # 候选集项Cn生成的频繁项集Lk
29     supportData = {}    #候选集项Cn的支持度字典
30     #计算候选项集的支持度, supportData key:候选项， value:支持度
31     for key in ssCnt:
32         support = ssCnt[key] / numItems
33         if support >= minSupport:
34             Lk.append(key)
35         supportData[key] = support
36     return Lk, supportData
37
38 #连接操作，将频繁Lk-1项集通过拼接转换为候选k项集
39 def aprioriGen(Lk_1, k):
40     Ck = []
41     lenLk = len(Lk_1)
42     for i in range(lenLk):
43         L1 = list(Lk_1[i])[:k - 2]
44         L1.sort()
45         for j in range(i + 1, lenLk):
46             #前k-2个项相同时，将两个集合合并
47             L2 = list(Lk_1[j])[:k - 2]
48             L2.sort()
49             if L1 == L2:
50                 Ck.append(Lk_1[i] | Lk_1[j])
51
52     return Ck
53
54 def apriori(dataSet, minSupport = 0.5):
55     C1 = createC1(dataSet)
56     L1, supportData = scanD(dataSet, C1, minSupport)
57     L = [L1]
58     k = 2
59     while (len(L[k-2]) > 0):
60         Lk_1 = L[k-2]
61         Ck = aprioriGen(Lk_1, k)
62         print("ck:",Ck)
63         Lk, supK = scanD(dataSet, Ck, minSupport)
64         supportData.update(supK)
65         print("lk:", Lk)
66         L.append(Lk)
67         k += 1
68
69     return L, supportData
70
71 dataset = loadDataSet()
72 L, supportData = apriori(dataset, minSupport=0.2)

 1 def scanD(D, Ck, minSupport):
 2     #候选集计数
 3     ssCnt = {}
 4     #数据集过滤
 5     D2 = [item for item in D if len(item) >= len(Ck[0])]
 6     for tid in D2:
 7         for can in Ck:
 8             if can.issubset(tid):
 9                 if can not in ssCnt.keys(): ssCnt[can] = 1
10                 else: ssCnt[can] += 1
11     ……
12
13     return Lk, supportData

#将频繁Lk-1项集转换为候选k项集
def aprioriGen(Lk_1, k):Ck = []lenLk = len(Lk_1)for i in range(lenLk):L1 = Lk_1[i]for j in range(i + 1, lenLk):L2 = Lk_1[j]if len(L1 & L2) == k - 2:L1_2 = L1 | L2if L1_2 not in Ck:Ck.append(L1 | L2)return Ck