Question

我有一个表，有三列。

id，时间和值

如何获得正确的ID，即每5分钟在Postgres中只有一个数据？

例如：

id    time    value 

1     03:11    test
1     07:13    test
1     12:11    test
2     02:11    test
2     07:11    test

这是我的版本，但只能得到一个小时的时间...

select
  distinct on("id")
  "id",
  time::time,
  value
from (
select
 "id",
 time::time,
 value,
dense_rank() over (
  partition by "id"
  order by
  to_char(time,'HH')::int,
  (
    to_char(time,'MI')
  )::int/5
)
from my_table
where time between '2018-07-03 00:00:00' and '2018-07-03 01:00:00' and "id" = 'XXXXX'
) as res
order by "id";

结果：

id   time  value
1     03:05   14

但是结果是有10行。.也许更多

Answer 1

下面的查询将产生所需的结果。

with res(id,time,value,rnk) as (
select
 id,
 time,
 value,
dense_rank() over (
  partition by id
  order by 
  to_char(time,'HH24'),
  (
    to_char(time,'MI')
  )::int/5
)
from your_table
)
select
  distinct on(id,rnk)
  id,
  to_char(time,'HH24:MI'),
  value
from res
order by id,rnk

如何在Postgres中每5分钟获得1行？

1 个答案: