Apriori算法python实现

以下是一个基于Python的Apriori算法实现：

# 导入必要的库
import itertools

# 定义函数，用于从数据集中获取所有项的列表
def get_items(data_set):
    items = set()
    for transaction in data_set:
        for item in transaction:
            items.add(frozenset([item]))
    return items

# 定义函数，用于从数据集中获取所有满足最小支持度要求的频繁项集
def get_frequent_items(data_set, min_support):
    items = get_items(data_set)
    frequent_items = []
    support_dict = {}
    for transaction in data_set:
        for item in items:
            if item.issubset(transaction):
                support_dict[item] = support_dict.get(item, 0) + 1
    for item, support in support_dict.items():
        if support / len(data_set) >= min_support:
            frequent_items.append(item)
    return frequent_items

# 定义函数，用于从频繁项集中获取所有满足最小置信度要求的关联规则
def get_association_rules(frequent_items, min_confidence):
    association_rules = []
    for item in frequent_items:
        if len(item) > 1:
            for i in range(1, len(item)):
                for subset in itertools.combinations(item, i):
                    antecedent = frozenset(subset)
                    consequent = item.difference(antecedent)
                    if antecedent in frequent_items and consequent in frequent_items:
                        confidence = support_dict[item] / support_dict[antecedent]
                        if confidence >= min_confidence:
                            association_rules.append((antecedent, consequent, confidence))
    return association_rules

以上代码实现了Apriori算法的核心功能，包括获取所有项的列表、获取所有满足最小支持度要求的频繁项集以及获取所有满足最小置信度要求的关联规则。具体使用时，可以将数据集转换为列表形式，并调用以上函数即可。