Redshift 查询以计算除他们创建的用户之外的用户观看的票数

Question

我们的票务系统中有 2 张桌子。在这个系统中，用户可以创建门票，任意数量的用户可以观看一张或多张门票。

为此，我们设置了 2 个表：

存储手表信息 - watch_table
存储所有权信息 - ownership_table

watch_table

ticket_id | watched_by
=======================
t1        | u1
t2        | u2
t2        | u1


ownership_table
ticket_id | owned_by
=====================
t1        | u1
t2        | u2

我想提取每个用户的观看次数。但是，不包括自己门票上的手表。

例如，在上面的例子中，输出应该是：

output:

user_id | count
================
u1      | 1
u2      | 0

我可以使用子查询创建查询来执行此操作。如下所示：

select watched_by as user_id, count(*) from watch_table wt where
(select count(*) from ownership_table where owned_by = wt.watched_by and ticket_id = wt.ticket_id) = 0
group by watched_by

我的问题是如何使用联接执行此操作，哪个性能更高？

Answer 1

根据您的示例数据，您可能可以使用 outer join

select t.owned_by as User_Id, Count(w.ticket_id) as count
from ownership_table t 
left join watch_table w on w.watched_by=t.owned_by and w.ticket_Id != t.ticket_Id
group by t.owned_by

替代方法可能是相关子查询：

select Owned_By as User_Id, (
    select Count(*)
    from watch_table w
    where w.watched_by=t.Owned_by and w.ticket_Id != t.ticket_Id
    )
from ownership_table t

Answer 2

select w.userid, count(o.ticketid)
from watches w left join owners o
    on o.ticketid = w.ticketid and o.userid <> w.userid
group by w.userid

https://dbfiddle.uk/?rdbms=sqlserver_2014&fiddle=19901f70822260cc943e6e51b42c7fbe

Redshift 查询以计算除他们创建的用户之外的用户观看的票数

Redshift query to count number of tickets watching by the users except the ones they created

sql

amazon-redshift