关于normalize函数record[i,:] != 0
来源：3-2 -电影推荐系统-构建模型
qq_三斤_0
2020-02-07 10:55
record[i,:] != 0这个能获取!=0的列下标吗，大家也偶人试了没，我这不行的
写回答
关注
1回答
-
- qq_三斤_0
- 2020-02-07 11:16:08
# 我改成了这样可以运行 for i in range(m): #获取一部电影评分用户的id ids = np.nonzero(record[i])[0] #ids = record[i,:] != 0 rating_mean[i] = np.mean(rating[i,ids]) #rating_norm[i,ids] -= rating_mean[i] rating_norm[i,ids] = rating[i,ids] - rating_mean[i] print("The row is {},mean rating is {},rating user size is {}" .format(i,rating_mean[i],ids.shape))
推荐算法理论与实践
想要了解商品推荐背后的原理吗，课程将手把手带你做一个推荐系统！
22159 学习 · 33 问题
相似问题