def match_dict_S1(item, user_bought_1_day_list, dict_item):
a = frequency_cat.collocation()
dict_utlimate = {}
dict_item_s1 = {}
for each in user_bought_1_day_list:
num_item = fpm(item, each, dict_item)
num_cat = a.cat_mat_frequency(item, each)
n = S1(num_item, num_cat)
dict_item_s1[each] = n
dict_utlimate[item] = dict_item_s1
print dict_utlimate
return dict_utlimate
def match_dict_S1(item, user_bought_1_day_list, dict_item):
'''
用户同日购买异类商品算法
'''
a = frequency_cat.collocation()
dict_utlimate = {}
dict_item_s1 = {}
for each in user_bought_1_day_list:
# fpm值计算
num_item = fpm(item, each, dict_item)
# 查询类间频率fcm
num_cat = a.cat_mat_frequency(item, each)
# 计算S1值
n = S1(num_item, num_cat)
dict_item_s1[each] = n
dict_utlimate[item] = dict_item_s1
print(dict_utlimate)
return dict_utlimate
np.dot(veca, vecb) / (np.linalg.norm(veca) * np.linalg.norm(vecb)))
return dist
def S2(fcm, cos_value):
'''
计算S2值
:param fcm:
:param cos_value:
:return:
'''
return cos_value * math.log((1 + fcm), math.e)
if __name__ == "__main__":
a = frequency_cat.collocation()
item_id = str(264)
tfidf_item = open('./data/tfidf_item.txt', 'r')
tfidf_item_all = tfidf_item.readlines()
tfidfList = tfidf_list(item_id, tfidf_item_all)
# print(tfidfList)
match_items = 0
for i in tfidf_item_all:
itemList = i.split("|")
goods_item = itemList[0]
if goods_item != item_id:
goodsTfidf = itemList[1]
goodsTfidf = goodsTfidf.replace("[", '')
goodsTfidf = goodsTfidf.replace("]", '')
goodsTfidf = goodsTfidf.replace(" ", '')
goodsTfidf = ast.literal_eval(goodsTfidf)