将数字列拆分为多列

Question

我有 SQL 查询，它从两个表中对所有者及其状态进行分组：

select value, count(distinct owner_id) as owners, account.status
from accounts_login_history
left join accounts on accounts.id = accounts_login_history.owner_id
where date >= '2020-02-01'
and owner_type = 1
group by value, accounts.status

输出：

我应该如何将拆分status列的查询更改为所有分类值（状态有 5 个唯一值）？

我使用 postgresql。

谢谢！

Answer 1

您可以使用条件聚合，在 Postgres 中最好使用filter来完成：

select value, count(distinct owner_id) as owners,
       max(a.status) filter (where status = 1) as status_1,
       max(a.status) filter (where status = 2) as status_2,
       max(a.status) filter (where status = 3) as status_3,
       max(a.status) filter (where status = 4) as status_4,
       max(a.status) filter (where status = 5) as status_5
from accounts_login_history alh left join
     accounts a 
     on accounts.id = alh.owner_id
where date >= '2020-02-01' and
      owner_type = 1
group by value;

这将为缺失值返回null而不是0 。 您当然可以使用coalesce()将其更改为0 。 但是，我更喜欢null 。