SQL计算库存成本

Question

我有两张桌子。一个持有特定单位的数量。另一个保存这些单位历史上到达时的数量、日期和成本。

所以在我的第一个表中，我会说 UNIT "ABC" 的数量为 50

在第二个表中，我的数据如下所示：

Unit   Date arrived   Quantity   Cost
----   ------------   --------   -----
ABC       11/1            100    $3.00
ABC       11/4             15    $5.00
ABC       11/5             25    $6.00

所以在本例中，我需要根据先进先出系统计算 50 件商品的价值。 50 个项目的数学运算如下所示：

25 x $6.00
15 x $5.00
10 x $3.00

该商品的总价值为 255.00 美元

所以我需要处理大约 300,000 个项目，并且我需要一个“简单按钮”。目前使用 MS Access & SQL 来挖掘我的数据。因此，与这些平台中的任何一个相关的任何解决方案都会很棒。

Answer 1

这个：-

select s.unit, dv.cost_2d+((s.quantity-dv.quantity_2d)*dv.cost) as valuation
from (
    select 
        d.*, 
        isnull((
            select sum(csq.quantity) 
            from #delivery csq 
            where csq.unit=d.unit 
                and csq.arrived>d.arrived
        ),0) as quantity_2d,
        isnull((
            select sum(csq.quantity*csq.cost) 
            from #delivery csq 
            where csq.unit=d.unit 
                and csq.arrived>d.arrived
        ),0) as cost_2d
    from #delivery d

    -- possible optimization - reduces the number of rows
    -- if we have enough to calculate value of stock held

    --join #stock s on s.unit=d.unit
    --and isnull((
    --    select sum(csq.quantity)
    --    from #delivery csq
    --    where csq.unit=d.unit 
    --        and csq.arrived>d.arrived
    --    ),0)<=s.quantity 

    -- you'd need to test if it helps/hinders with your dataset/schema

) as dv
join #stock s on s.unit=dv.unit
where dv.quantity+dv.quantity_2d>=s.quantity 
    and dv.quantity_2d<s.quantity

产生：-

unit    valuation
abc     255.00

如果来自：-

create table #stock (
    unit varchar(10),
    quantity int
)

create table #delivery (
    unit varchar(10),
    arrived date,
    quantity int,
    cost money
)

insert into #stock values ('abc',50)
insert into #delivery values ('abc','2013-11-01',100,3)
insert into #delivery values ('abc','2013-11-04',15,5)
insert into #delivery values ('abc','2013-11-05',25,6)

-----------更新-----------------------------------

这是另一个可能会或可能不会运行得更快的版本 - 取决于您的数据集/架构：-

select dv.unit, dv.cost_2d+((dv.instock-dv.quantity_2d)*dv.cost) as valuation
from (
    select 
        d.*, 
        isnull((
            select sum(csq.quantity) 
            from #delivery csq 
            where csq.unit=d.unit 
                and csq.arrived>d.arrived
        ),0) as quantity_2d,
        isnull((
            select sum(csq.quantity*csq.cost) 
            from #delivery csq 
            where csq.unit=d.unit 
                and csq.arrived>d.arrived
        ),0) as cost_2d,
        s.quantity as instock
    from #delivery d
    join #stock s on s.unit=d.unit
    and s.quantity between isnull((
            select sum(csq.quantity) 
            from #delivery csq 
            where csq.unit=d.unit 
                and csq.arrived>d.arrived
        ),0) and isnull((
            select sum(csq.quantity) 
            from #delivery csq 
            where csq.unit=d.unit 
                and csq.arrived>d.arrived
        ),0) + d.quantity
) as dv