def user_features():
"""用户特征"""
users_feature = {}
for line in loadfile(base_path + "ml-1m/users.dat", encoding="ISO-8859-1"):
arr = line.split("::")
sex_feature = judge_sex(arr[1])
age_feature = judge_age(int(arr[2]))
user_feature_list = sex_feature + age_feature + [int(arr[3])]
users_feature.setdefault(arr[0], user_feature_list)
return users_feature
def user_item_rating():
"""user-item"""
user_item = {}
for line in loadfile(base_path + "ml-1m/ratings.dat"):
arr = line.split("::")
uid = arr[0]
user_item.setdefault(uid, [[], []])
user_item[uid][0].append(arr[1])
user_item[uid][1].append(arr[2])
return user_item
def create_user_item_dict():
"""生成user-item 字典"""
item_dict = create_item_dict()
user_item_dict = {}
for line in loadfile(base_path + "ml-1m/ratings.dat",
encoding="ISO-8859-1"):
arr = line.split("::")
user_item_dict.setdefault(arr[0], list())
movie = item_dict.get(arr[1])
user_item_dict.get(arr[0]).append(movie)
return user_item_dict
def load_data(train_rate=1):
train_items_ratings = {}
test_data = {}
for line in loadfile(base_path + "ml-1m/ratings.dat"):
arr = line.split("::")
if random.random() < train_rate:
train_items_ratings.setdefault(arr[1], set())
train_items_ratings[arr[1]].add(arr[2])
else:
test_data.setdefault(arr[0], set())
test_data[arr[0]].add(arr[1])
return train_items_ratings, test_data
def create_item_dict():
"""生成item 字典"""
item_dict = {}
for line in loadfile(base_path + "ml-1m/movies.dat",
encoding="ISO-8859-1"):
arr = line.split("::")
item_dict.setdefault(arr[0], {})
title = remove_punctuation(arr[1])
genres = ' '.join(g.strip('\n') for g in arr[2].split("|"))
item_dict[arr[0]]['title'] = title
item_dict[arr[0]]['genres'] = genres
return item_dict
def item_features():
"""商品特征矩阵"""
item_feature = {}
for line in loadfile(base_path + "ml-1m/movies.dat",
encoding="ISO-8859-1"):
arr = line.split("::")
year = int(re.findall("\d{4}", arr[1])[0])
year_label = judge_year(year)
genres = arr[2].split("|")
genres_list = judge_genres(genres)
item_feature_list = year_label + genres_list
item_feature.setdefault(arr[0], item_feature_list)
return item_feature
item_feature_list = year_label + genres_list
item_feature.setdefault(arr[0], item_feature_list)
return item_feature
if __name__ == '__main__':
count_2000 = 0
count_1990 = 0
count_1980 = 0
count_1970 = 0
other = 0
item_feature = {}
for line in loadfile(base_path + "ml-1m/movies.dat",
encoding="ISO-8859-1"):
arr = line.split("::")
year = int(re.findall("\d{4}", arr[1])[0])
year_label = judge_year(year)
genres = arr[2].split("|")
genres_list = judge_genres(genres)
item_feature_list = year_label + genres_list
item_feature.setdefault(arr[0], item_feature_list)
if year >= 2000:
count_2000 += 1
elif year >= 1990 and year < 2000:
count_1990 += 1
elif year >= 1980 and year < 1990:
count_1980 += 1
elif year >= 1970 and year < 1980:
count_1970 += 1