Apriori实现后封装的函数如下,其中support代表支持度,minConf代表置信度:
L, suppData = Apriori(myDat, support=0.5) rules = generateRules(L, suppData, minConf=0.5)
假设我们需要从下列数据中挖掘频繁项集:
myDat = [ [ 1, 3, 4 ], [ 2, 3, 5 ], [ 1, 2, 3, 5 ], [ 2, 5 ] ]
满足的条件为支持度为0.5,置信度为0.7:
L, suppData = apriori(myDat, 0.5) rules = generateRules(L, suppData, minConf=0.7) print 'rules:\n', rules
输出结果为:
frozenset([1]) --> frozenset([3]) conf: 1.0 frozenset([5]) --> frozenset([2]) conf: 1.0 frozenset([2]) --> frozenset([5]) conf: 1.0 rules: [(frozenset([1]), frozenset([3]), 1.0), (frozenset([5]), frozenset([2]), 1.0), (frozenset([2]), frozenset([5]), 1.0)]