import numpyas np
import pandasas pd
#1、读入数据集
header = ["user_id","item_id","rating","timestamp"]
data =pd.read_csv("u.data",sep="\t",names=header)
#2、生成用户—物品评分矩阵
#检查是否有重复的用户物品打分记录
data.duplicated(subset = ["user_id","item_id"]).sum()
item_id_user = data.groupby("item_id").count()["user_id"]
#构建用户物品矩阵
users_num = data.user_id.max()
items_num = data.item_id.max()
user_item_rating = np.zeros((users_num,items_num))
for linein data.itertuples():#以元组的方式赋值
user_item_rating[line[1]-1,line[2]-1] = line[3]
np.savetxt("评分12.csv", user_item_rating,delimiter =",")
