语法分组依据

Question

我需要计算性能，即 7 月份的最大里程减去 8 月份的最大里程，试试这个有效但只打印相同数量的查询。

这是我的查询：

select vehicle_id, max(odometro) as cierre,sum(cantidad) as lts,((select max(odometro) 
      from vehicle_consumptions where fecha between '2020-07-01' and '2020-07-31' group by vehicle_id) - (select max(odometro) from vehicle_consumptions where fecha between '2020-08-01' and '2020-08-30' group by vehicle_id))
       as recorrido,
       ((select max(odometro) 
      from vehicle_consumptions where fecha between '2020-07-01' and '2020-07-31') - 
      (select max(odometro) from vehicle_consumptions where fecha between '2020-08-01' and '2020-08-30'))/sum(cantidad) as rendimiento
        from vehicle_consumptions where EXTRACT(month from fecha)= 08 group by vehicle_id

Answer 1

我想你想要条件聚合：

select vehicle_id,
    max(case when fecha < '2020-08-01' then odometro end)
        - max(case when fecha >= '2020-08-01' then odometro end)
    as odometro_diff
from vehicle_consumptions 
where fecha between '2020-07-01' and '2020-08-31'
group by vehicle_id

在 Postgres 中，我们可以使用标准的 filter 子句和日期文字（假设 fecha 是一种类似日期的数据类型——它应该是）：

select vehicle_id,
    max(odometro) filter(fecha < date '2020-08-01')
        - max(odometro) filter(where fecha >= date '2020-08-01')
    as odometro_diff
from vehicle_consumptions 
where fecha between date '2020-07-01' and date '2020-08-31'
group by vehicle_id

最后：如果你的日期有时间成分，我会推荐半开间隔，所以 where 子句应该是：

where fecha >= date '2020-07-01' and fecha < '2020-09-01'

Answer 2

您可以使用条件聚合：

select vehicle_id,
       (max(case when fecha between '2020-08-01' and '2020-08-31' then odometro end)
        max(case when fecha between '2020-07-01' and '2020-07-31' then odometro end)
       ) as diff 
from vehicle_consumption
group by vehicle_id;

在 Postgres 中，您可以将其表述为：

select vehicle_id,
       (max(odometro) filter (where extract(month from fecha) = 7) -
        max(odometro) filter (where extract(month) from fecha) = 8)
       ) as diff 
from vehicle_consumption
where fecha >= '2020-07-01' and fecha < '2020-09-01'
group by vehicle_id;

where 子句将处理时间限制为仅两个月，这应该会使查询执行得更好。然后它只使用月份来过滤 max().

语法分组依据

syntax group by

sql

postgresql

pivot

max

aggregate-functions