python实现简单关联规则Apriori算法

 from itertools import combinations
 from copy import deepcopy

 # 导入数据，并剔除支持度计数小于min_support的1项集
 def load_data(data):
     I_dict = {}
     for i in data:
         for j in i:
             I_dict[j] = I_dict.get(j, 0) + 1
     F_dict = deepcopy(I_dict)
     for k in I_dict.keys():
         if F_dict.get(k) < min_support:
             del F_dict[k]
     return F_dict

 # 判断频繁项集是否大于min_support
 def get_support_set(p_set):
     item_supp_set = []
     for item in p_set:
         count = 0
         for ds in data_set:
             if item.issubset(ds):
                 count += 1
         if count >= min_support:
             item_supp_set.append([item, count])
     return item_supp_set

 # 找出所有频繁项集
 # 以二项集为初始集
 def get_all_items(two_set, k=3):
     all_frequent = []
     flag = True
     while flag:
         mid_set = []
         temp = []
         t_ = [ks[0] for ks in two_set]
         for kk in t_:
             for tt in kk:
                 if tt not in temp:
                     temp.append(tt)
         k_ = [set(t) for t in combinations(temp, k)]
         for ff in k_:
             count_k = 0
             for d in t_:
                 if ff.issuperset(d):
                     count_k += 1
             if count_k == k:
                 mid_set.append(ff)
         frequent_mid_set = get_support_set(mid_set)
         if mid_set:
             k += 1
             two_set = frequent_mid_set
             all_frequent.extend(frequent_mid_set)
         else:
             flag = False
     return all_frequent

 if __name__ == '__main__':
     data = [['I1', 'I2', 'I5'],
             ['I2', 'I4'],
             ['I2', 'I3'],
             ['I1', 'I2', 'I4'],
             ['I1', 'I3'],
             ['I2', 'I3'],
             ['I1', 'I3'],
             ['I1', 'I2', 'I3', 'I5'],
             ['I1', 'I2', 'I3']]
     data_set = [set(d) for d in data]
     min_support = 1
     one = [[{lk}, lv] for lk, lv in load_data(data).items()]
     two = [set(t) for t in combinations(list(load_data(data).keys()), 2)]
     two_f_set = get_support_set(two)
     all_frequent_set = one + two_f_set + get_all_items(two_f_set)
     for afs in all_frequent_set:
         print(afs)

输出结果：