MSapriori和CARapriori算法在Python或R中的实现

def msapriori(data, min_support):
    # 初始化频繁1项集
    freq_1_itemsets = {frozenset([item]) for item in set(item for transaction in data for item in transaction)}
    
    # 过滤出满足最小支持度的频繁1项集
    freq_1_itemsets = {itemset for itemset in freq_1_itemsets if support(itemset, data) >= min_support}
    
    # 初始化频繁k项集列表
    freq_itemsets = [freq_1_itemsets]
    
    k = 2
    while freq_1_itemsets:
        # 生成候选k项集
        candidate_k_itemsets = apriori_gen(freq_1_itemsets, k)
        
        # 计算候选k项集的支持度
        freq_k_itemsets = {itemset for itemset in candidate_k_itemsets if support(itemset, data) >= min_support}
        
        # 更新频繁k项集列表
        freq_itemsets.append(freq_k_itemsets)
        
        # 更新频繁1项集为下一轮迭代的候选k项集
        freq_1_itemsets = freq_k_itemsets
        
        k += 1
    
    return freq_itemsets

def apriori_gen(freq_itemsets, k):
    # 生成候选k项集
    candidate_k_itemsets = set()
    for itemset1 in freq_itemsets[-1]:
        for itemset2 in freq_itemsets[-1]:
            union = itemset1.union(itemset2)
            if len(union) == k:
                candidate_k_itemsets.add(union)
    return candidate_k_itemsets

def support(itemset, data):
    # 计算项集的支持度
    count = 0
    for transaction in data:
        if itemset.issubset(transaction):
            count +=  0
    return count / len(data)

# 示例数据
data = [
    ['牛奶', '面包', '黄油'],
    ['啤酒', '面包', '尿布'],
    ['牛奶', '尿布', '饼干'],
    ['啤酒', '饼干'],
    ['牛奶', '面包', '尿布', '啤酒'],
    ['牛奶', '尿布', '饼干', '面包']
]

# 最小支持度
min_support = 0.5

# 运行MSapriori算法
freq_itemsets = msapriori(data, min_support)
print(freq_itemsets)

def carapriori(data, min_support):
    # 初始化频繁1项集
    freq_1_itemsets = {frozenset([item]) for item in set(item for transaction in data for item in transaction)}
    
    # 过滤出满足最小支持度的频繁1项集
    freq_1_itemsets = {itemset for itemset in freq_1_itemsets if support(itemset, data) >= min_support}
    
    # 初始化频繁k项集列表
    freq_itemsets = [freq_1_itemsets]
    
    k = 2
    while freq_1_itemsets:
        # 生成候选k项集
        candidate_k_itemsets = carapriori_gen(freq_1_itemsets, k)
        
        # 计算候选k项集的支持度
        freq_k_itemsets = {itemset for itemset in candidate_k_itemsets if support(itemset, data) >= min_support}
        
        # 更新频繁k项集列表
        freq_itemsets.append(freq_k_itemsets)
        
        # 更新频繁1项集为下一轮迭代的候选k项集
        freq_1_itemsets = freq_k_itemsets
        
        k += 1
    
    return freq_itemsets

def carapriori_gen(freq_itemsets, k):
    # 生成候选k项集
    candidate_k_itemsets = set()
    for itemset1 in freq_itemsets[-1]:
        for itemset2 in freq_itemsets[-1]:
            union = itemset1.union(itemset2)
            if len(union) == k:
                candidate_k_itemsets.add(union)
    return candidate_k_itemsets

def support(itemset, data):
    # 计算项集的支持度
    count = 0
    for transaction in data:
        if itemset.issubset(transaction):
            count += 1
    return count / len(data)

# 示例数据
data = [
    ['牛奶', '面包', '黄油'],
    ['啤酒', '面包', '尿布'],
    ['牛奶', '尿布', '饼干'],
    ['啤酒', '饼干'],
    ['牛奶', '面包', '尿布', '啤酒'],
    ['牛奶', '尿布', '饼干', '面包']
]

# 最小支持度
min_support = 0.5

# 运行CARapriori算法
freq_itemsets = carapriori(data, min_support)
print(freq_itemsets)

# 安装并加载arules包
install.packages("arules")
library(arules)

# 示例数据
data <- list(
  c("牛奶", "面包", "黄油"),
  c("啤酒", "面包", "尿布"),
  c("牛奶", "尿布", "饼干"),
  c("啤酒", "饼干"),
  c("牛奶", "面包", "尿布", "啤酒"),
  c("牛奶", "尿布", "饼干", "面包")
)

# 转换为transactions对象
trans <- as(data, "transactions")

# 运行MSapriori算法
rules <- apriori(trans, parameter = list(support = 0.5))
print(rules)

# 运行CARapriori算法
rules_car <- apriori(trans, parameter = list(support = 0.5, maxlen = 3))
print(rules_car)