关联分析购物篮案例python_关联分析(5):Apriori Python实现

import pandas as pd

import numpy as np

import operator

from efficient_apriori import apriori

#读取数据

data = pd.read_csv("voting-records.csv",header=None)

#得到频繁项集及关联规则

itemsets, rules = apriori(data.values.tolist(), min_support=0.3,  min_confidence=0.9,max_length=10)

#分别根据置信度、支持度、提升度对规则进行排序

confidence = dict()

support = dict()

lift = dict()

for ruleinrules:

confidence[rule] = rule.confidence

support[rule] = rule.support

lift[(rule)] = rule.lift

#rule.lhs,rule.rhs可用作提取规则的前件、后件

rules_sortbycon = sorted(confidence.items(),key=operator.itemgetter(1),reverse = True)

rules_sortbysup = sorted(support.items(),key=operator.itemgetter(1),reverse = True)

rules_sortbylift = sorted(lift.items(),key=operator.itemgetter(1),reverse = True)