SQL：如何对 Sum 列求和？

Posted 2023-03-21

【中文标题】SQL：如何对 Sum 列求和？

【英文标题】：SQL: How do you Sum a Sum column?

【发布时间】：2021-08-31 12:32:53

【问题描述】：

我的查询返回我需要的列，但我想知道是否有办法让另一列返回总和： sum(CAST(qty AS int)) 作为 qty 列。如果每行都列出它就可以了；我只是想要一种方法，不必拿出计算器并将所有的总和相加（CAST（qty AS int））。

select sql1.item_code, sql1.supplier_so, sum(sql1.qty) Qty, sql1.dt
from
(SELECT supplier_so, item_code, sum(CAST(qty AS int)) as qty, dt FROM "table.name"
where
supplier_so = 'XXXXXXX'
group by supplier_so, item_code, dt
UNION
SELECT supplier_so, item_code, sum(CAST(qty AS int)) as qty, dt FROM "table.name2"
where
supplier_so = 'XXXXXXX'
and
dt < date '2021-06-01'
 group by supplier_so, item_code, dt
) sql1
group by supplier_so, item_code, dt

【问题讨论】：

检查您的 RDBMS 是否支持ROLLUP 功能。

看起来确实如此，但它所做的只是复制每一行

【参考方案1】：

您可以为此使用window function：

SELECT 
   ..., 
   SUM(CAST(qty AS int)) OVER ()
FROM ...

【参考方案2】：

你为什么不能试试这个。

with sql1 as 
(SELECT supplier_so, item_code, sum(CAST(qty AS int)) as qty, dt FROM "table.name"
where
supplier_so = 'XXXXXXX'
group by supplier_so, item_code, dt
UNION
SELECT supplier_so, item_code, sum(CAST(qty AS int)) as qty, dt FROM "table.name2"
where
supplier_so = 'XXXXXXX'
and
dt < date '2021-06-01'
 group by supplier_so, item_code, dt
) 
select sql1.item_code, sql1.supplier_so, sql1.dt, sum(sql1.qty) over () from sql1;

【讨论】：

所以这确实给了我想要的专栏，谢谢！但是，它去掉了原始的 qty 列，该列显示所有单独的结果，这些结果加起来是窗口化的 sum(sql1.qty) over ()

如果你想在输出中也有 qty，你可以在外部查询中添加。它应该工作。 select sql1.item_code, sql1.supplier_so, sql1.dt, sql1.qty, sum(sql1.qty) over () from sql1;