SQL查询以在一段时间内每天选择最小值和最大值

时间:2019-05-15 20:58:43

标签: sql hive hiveql

我想select包含每个equipment_id的最小日期时间值或最大日期时间值的所有行,并持续一段时间(每天)。

下面的代码在整个期间内为每个min选择maxequipment_id日期时间值。我可以/应该改变什么以达到期望的选择结果?

注意:代码段中用大括号括起来的值表示动态值。

select *
from equipment
where created_at in (select min(created_at)
                     from equipment
                     where created_at >= {start_datetime} and created_at < {end_datetime}
                     group by equipment_id
                    ) or
      created_at in (select max(created_at)
                     from equipment
                     where created_at >= {start_datetime} and created_at < {end_datetime}
                     group by equipment_id
                    )
order by account, equipment_id, created_at asc;

2 个答案:

答案 0 :(得分:1)

使用窗口功能:

select e.*
from (select e.*,
             min(e.created_at) over (partition by e.equipment_id, to_date(e.created_at)) as min_ca,
             max(e.created_at) over (partition by e.equipment_id, to_date(e.created_at)) as max_ca
      from equipment e
     ) e
where e.created_at in (min_ca, max_ca)
order by account, equipment_id, created_at asc;

答案 1 :(得分:0)

使用to_date从datetime值中提取日期。然后,加入汇总查询:

with agg as
  (select to_date(created_at) as created_date, 
          equipment_id,
          min(created_at) as min_created_at,
          max(created_at) as max_created_at
   from equipment
   group by to_date(created_at),
            equipment_id
  )

select e.*
from equipment e
left join agg
   on e.equipment_id = agg.equipment_id
   and to_date(e.created_at) = agg.created_date

where e.created_at in (agg.min_created_at, 
                       agg.max_created_at)