I have two tables user (one)
and transaction (many)
and I need to get the average time in days from when a user was created to when they made their first transaction. I'm using AVG(TIMESTAMPDIFF)
which is working well, except that the GROUP BY
returns an average against every user instead of one single average for all unique users in the transaction
table. If I remove the GROUP BY
, I get a single average figure but it takes into account multiple trans开发者_如何转开发actions from users, whereas I just want to have one per user (the first they made).
Here's my SQL:
SELECT AVG(TIMESTAMPDIFF(DAY, u.date_created, t.transaction_date)) AS average
FROM transaction t
LEFT JOIN user u ON u.id = t.user_id
WHERE t.user_id IS NOT NULL AND t.status = 1
GROUP BY t.user_id;
I'd appreciate it if someone can help me return the average for unique users only. It's fine to break the query down into two, but the tables are large so returning lots of data and putting it back in is a no-go. Thanks in advance.
SELECT AVG(TIMESTAMPDIFF(DAY, S.date_created, S.transaction_date)) AS average
FROM (
SELECT u.date_created, t.transaction_date
FROM transaction t
INNER JOIN user u ON u.id = t.user_id
WHERE t.status = 1
GROUP BY t.user_id
HAVING u.date_created = MIN(u.date_created)
) s
I replaced the LEFT JOIN with an INNER JOIN because I think that's what you want, but it's not 100% equivalant to your WHERE t.user_id IS NOT NULL
.
Feel free to put the LEFT JOIN back if need be.
select avg( TIMESTAMPDIFF(DAY, u.date_created, min_tdate) ) as average
from user u
inner join
(select t.user_id, min(t.transaction_date) as min_tdate
from transaction t
where t.status=1;
group by t.user_id
) as min_t
on u.id=min_t.user_id;
精彩评论