用python实现Apriori 关联规则

Apriori算法是一种用于挖掘频繁项集和关联规则的经典算法。下面是用Python实现Apriori算法的步骤：

1. 导入必要的库

import itertools
from collections import defaultdict

2. 定义函数 `get_frequent_itemsets`，用于获取频繁项集

def get_frequent_itemsets(transactions, min_support):
    item_counts = defaultdict(int)
    for transaction in transactions:
        for item in transaction:
            item_counts[item] += 1

    num_items = len(item_counts)
    min_count = min_support * len(transactions)
    frequent_items = set(item for item, count in item_counts.items() if count >= min_count)

    if len(frequent_items) == 0:
        return set()

    frequent_itemsets = set(frozenset([item]) for item in frequent_items)

    k = 2

    while True:
        new_frequent_itemsets = set()
        for itemset in itertools.combinations(frequent_itemsets, 2):
            union = itemset[0].union(itemset[1])
            if len(union) == k and union not in new_frequent_itemsets:
                count = sum(1 for transaction in transactions if union.issubset(transaction))
                if count >= min_count:
                    new_frequent_itemsets.add(union)

        if len(new_frequent_itemsets) == 0:
            break

        frequent_itemsets = new_frequent_itemsets
        k += 1

    return frequent_itemsets

3. 定义函数 `get_association_rules`，用于获取关联规则

def get_association_rules(frequent_itemsets, min_confidence):
    rules = []
    for itemset in frequent_itemsets:
        if len(itemset) > 1:
            for item in itemset:
                antecedent = itemset - set([item])
                consequent = set([item])
                confidence = float(len(itemset)) * sum(1 for transaction in transactions if itemset.issubset(transaction)) / sum(1 for transaction in transactions if antecedent.issubset(transaction))
                if confidence >= min_confidence:
                    rules.append((antecedent, consequent, confidence))
    return rules

4. 定义数据集和参数

transactions = [
    set(['A', 'B', 'C', 'D']),
    set(['A', 'B', 'C']),
    set(['A', 'B', 'D']),
    set(['A', 'B']),
    set(['A', 'C', 'D']),
    set(['A', 'C']),
    set(['A', 'D']),
    set(['A'])
]

min_support = 0.5
min_confidence = 0.6

5. 获取频繁项集和关联规则

frequent_itemsets = get_frequent_itemsets(transactions, min_support)
association_rules = get_association_rules(frequent_itemsets, min_confidence)

6. 打印结果

print('Frequent itemsets:')
for itemset in frequent_itemsets:
    print(itemset)

print('\nAssociation rules:')
for rule in association_rules:
    print(rule[0], '->', rule[1], ':', rule[2])

完整代码如下：