快手电商数据分析凉经
自己表达太差了所以一面挂吧。。。
问用户流失相关的问题,比如说应该给哪些人发放优惠券来确保他们不流失,怎么确保这类人不流失
SQL:
一个用户直播打赏表 user_author_live_amt,记录每天用户给每个主播打赏的金额,包含字段(p_date, user_id, author_id, total_cost_amt),分别为:格式为yyyymmdd的日期,打赏用户ID,被打赏主播ID,打赏金额)
问题:查询累计打赏超过10万元的用户中,每个用户打赏总金额排名前3的主播ID和其被打赏总金额
我的答案是:
with total as(
select
user_id,
sum(total_cost_amt) as total_sum
from user_author_live_amt
group by user_id
having total_sum > 1000000),
author as (
select
user_id,
author_id,
sum(total_cost_amt) as author_total
from user_author_live_amt
group by user_id,author_id
)
select user_id,
author_id, author_total
from (
select
a.user_id,a.author_id,a.author_total
row_number()over(partition by a.user_id order by a.author_total desc) ranking
from author a
join total t on a.user_id = t.user_id
)s
where ranking <= 3;
问用户流失相关的问题,比如说应该给哪些人发放优惠券来确保他们不流失,怎么确保这类人不流失
SQL:
一个用户直播打赏表 user_author_live_amt,记录每天用户给每个主播打赏的金额,包含字段(p_date, user_id, author_id, total_cost_amt),分别为:格式为yyyymmdd的日期,打赏用户ID,被打赏主播ID,打赏金额)
问题:查询累计打赏超过10万元的用户中,每个用户打赏总金额排名前3的主播ID和其被打赏总金额
我的答案是:
with total as(
select
user_id,
sum(total_cost_amt) as total_sum
from user_author_live_amt
group by user_id
having total_sum > 1000000),
author as (
select
user_id,
author_id,
sum(total_cost_amt) as author_total
from user_author_live_amt
group by user_id,author_id
)
select user_id,
author_id, author_total
from (
select
a.user_id,a.author_id,a.author_total
row_number()over(partition by a.user_id order by a.author_total desc) ranking
from author a
join total t on a.user_id = t.user_id
)s
where ranking <= 3;
全部评论
纯讨论一下sql是不是写的有点问题,total表having 用新命名字段筛不出来,第一张表找出用户之后直接join
WITH t0 AS (
SELECT
user_id
FROM user_author_live_amt
GROUP BY user_id
HAVING SUM(total_cost_amt) > 100000
),
t1 AS (
SELECT
a.user_id,
a.author_id,
SUM(a.total_cost_amt) AS total_reward,
ROW_NUMBER() OVER ( PARTITION BY a.user_id ORDER BY SUM(a.total_cost_amt) DESC) AS ranking
FROM user_author_live_amt a
JOIN t0 b ON a.user_id = b.user_id
GROUP BY 1, 2
)
SELECT
user_id,
author_id,
total_reward
FROM user_author_ranking
WHERE ranking <= 3
ORDER BY user_id, ranking;
相关推荐
07-19 09:44
四川师范大学 数据分析师 点赞 评论 收藏
分享