更改Mysql SQL命令的输出样式分组

如何解决更改Mysql SQL命令的输出样式分组

如何在此sql命令上创建过滤器视图？

select t1.ean,t2.date,t2.hour,(t1.consumption *
    (case when type = 'B' then type_b_coefficient
          when type = 'A' and region = 'EAST' then type_a_east_coefficient
          when type = 'A' and region = 'WEST' then type_a_west_coefficient
          else 1
     end) *
    (case when category = 11 then cat_11_coefficient
          when category = 21 then cat_21_coefficient
          else 1
     end)
   ) as net_consumption
from table1 t1 join
 table2 t2
 on t2.date = t1.from_date

现在查询显示EAN /小时/天/消耗量。我需要表演 EAN /天 天 以及DAY和Hour。

如果我通过查询更改GROUP会得出错误的数学结果。在此命令中，所有数学运算都很好

过滤器含义：

现在查询显示如下输出：

  ean      day      hour         net_consumption
  1234  2020-01-01   1              1.12312312
  1234  2020-01-01   2              0.23038382
  2838  2020-01-01   1              2.3043043

以及如何处理此视图-

 ean      day           net_consumption
1234  2020-01-01          10.12312312
5678  2020-01-01           8.12312312
2838  2020-01-01           5.22392229

或-

     day           net_consumption
   2020-01-01          120.12312312
   2020-01-02           68.12312312
   2020-01-03           30.22392229

如果我使用GROUP BY执行此操作，则会得出错误的数学结果（例如，如果我手动进行计数，则错误的计数消耗会低得多。第一个查询有效）

解决方法

参与JOIN的汇总可能比预期的大。

这就是发生的事情。

1. 将这些表结合起来，以[逻辑地]创建具有适当行组合的临时表。
2. 聚合是基于此[可能更大]的临时表计算的。
3. 然后GROUP BY缩小行数。

通常，解决数字膨胀（以及性能下降）的方法是首先用尽可能少的列和尽可能少的表进行聚合和分组。然后做

 SELECT ...
     FROM ( that query to get aggregates ) AS x
     JOIN table(s) ON ...  -- to get other columns

,

您应该尝试以下类似操作

Select 
    ean,date,Sum(net_consumption) as net_consumption
from
(
    select 
        t1.ean,t2.date,t2.hour,t1.consumption *
        case when type = 'B' then type_b_coefficient
              when type = 'A' and region = 'EAST' then type_a_east_coefficient
              when type = 'A' and region = 'WEST' then type_a_west_coefficient
              else 1
         end *
        case when category = 11 then cat_11_coefficient
              when category = 21 then cat_21_coefficient
              else 1
         end
        as net_consumption
    from table1 t1 join
     table2 t2
     on t2.date = t1.from_date

)T
group by ean,date

或

Select 
    date,t1.consumption *
        case when type = 'B' then type_b_coefficient
              when type = 'A' and region = 'EAST' then type_a_east_coefficient
              when type = 'A' and region = 'WEST' then type_a_west_coefficient
              else 1
         end *
        case when category = 11 then cat_11_coefficient
              when category = 21 then cat_21_coefficient
              else 1
         end
        as net_consumption
    from table1 t1 join
     table2 t2
     on t2.date = t1.from_date

)T
group by date