GROUP BY使用另一个表

Question

我有两个表用户和投票，有些用户可以拥有相同的unq_id

我尝试从用户中选择所有不同 votes，然后从投票 table。

但结果不是我的预期，投票是重复的。您可以在下图中看到

SELECT DISTINCT u.unq_id, COUNT(v.id) AS count_votes
FROM users u
INNER JOIN votes v
ON u.unq_id = v.unq_id
WHERE u.unq_id <> ''
GROUP BY u.unq_id;

SQLFiddle：

预期结果：

3fyx6 - 3
9kx9mq - 1

Answer 1

尝试更改

SELECT DISTINCT u.unq_id, COUNT(v.id) AS count_votes

到

SELECT DISTINCT u.unq_id, COUNT(distinct v.id) AS count_votes

看看是否有帮助：）

Answer 2

要获得所需的结果，您无需加入users，但根据您的评论“某些用户根本没有投票，我想将它们列入列表”

SELECT DISTINCT
   u.unq_id, 
   COALESCE(v.count_votes, 0)
FROM users u
LEFT JOIN
 ( select unq_id, count(*) as count_votes
   from votes
   group by unq_id
 ) v
ON u.unq_id = v.unq_id
WHERE u.unq_id <> '';

Answer 3

这样可行：

select distinct v.unq_id, count(distinct v.id) as votes
from users u
inner join votes v
on u.unq_id = v.unq_id
group by v.unq_id
order by 2 desc;

但我宁愿使用子查询，我认为它会更快：

select v.unq_id, count(v.unq_id) as votes
from votes v
where v.unq_id in (select unq_id from users group by unq_id)
group by v.unq_id
order by 2 desc;