apriori_rules.py

Apriori_rules.py

def generateRules(L, support_data, min_confidence=0.7):
    """Create the association rules
    L: list of frequent item sets
    support_data: support data for those itemsets
    min_confidence: minimum confidence threshold
    """
    rules = []
    for i in range(1, len(L)):
        for freqSet in L[i]:
            H1 = [frozenset([item]) for item in freqSet]
            print "freqSet", freqSet, 'H1', H1
            if (i > 1):
                rules_from_conseq(freqSet, H1, support_data, rules, min_confidence)
            else:
                calc_confidence(freqSet, H1, support_data, rules, min_confidence)
    return rules


def calc_confidence(freqSet, H, support_data, rules, min_confidence=0.7):
    "Evaluate the rule generated"
    pruned_H = []
    for conseq in H:
        conf = support_data[freqSet] / support_data[freqSet - conseq]
        if conf >= min_confidence:
            print freqSet - conseq, '--->', conseq, 'conf:', conf
            rules.append((freqSet - conseq, conseq, conf))
            pruned_H.append(conseq)
    return pruned_H


def rules_from_conseq(freqSet, H, support_data, rules, min_confidence=0.7):
    "Generate a set of candidate rules"
    m = len(H[0])
    if (len(freqSet) > (m + 1)):
        Hmp1 = aprioriGen(H, m + 1)
        Hmp1 = calc_confidence(freqSet, Hmp1,  support_data, rules, min_confidence)
        if len(Hmp1) > 1:
            rules_from_conseq(freqSet, Hmp1, support_data, rules, min_confidence)