Question

I am developing an Inventory system, using average cost method, where average cost price of product changes on each new purchase. Database I am using is SQL Server 2008

Now I need to calculate COST OF GOODS SOLD in a report, where I have to Add Cost Price for Each Sale of a product, but this COST PRICE should be relevant to a that specific Purchase period.

My Purchase Table

Purchase_Date   Product_ID   Ave_Cost_Price
1-jan-2013      1                  5.5
15-jan-2013     1                  6.5
30-jan-2013     1                  7.5

My Sales Table

Sale_Date   Product_ID    Sale_Price Cost_Price
5-jan-2013      1              10         ?         SALE-1
17-jan-2013     1              10         ?         SALE-1
31-jan-2013     1              15         ?         SALE-1

Now when I create a SALE report, SALE-1 should take 5.5, SALE-2, 6.5 and SALE-3 should pick 7.5 as Product cost price. and if it cannot find any purchase then it should pick opening Cost_Price from product_table.

I am looking for Such a Query which should do the job???

I think may be it can be done by some grouping and inner joins, but cannot figured it out.

any suggestions please ????

regards Raza

Answer 1

I’m not a big fan of correlated subqueries in the select clause. Part of the reason is aesthetic and mmaintainability. I prefer to have all table references in one place, in the from clause. Part of the reason is performance; I tend to think that they turn into nested loop joins (usually for worse). Admittedly, SQL optimizers have gotten much better in the last few years.

But this is one case where I think it is the best approach:

[trouble uploading]

The correlated subquery is . . .

select s.*,

           (select top 1 ave_cost_price
from purchase p
            where p.purchase_date <= sys.sale_date and p.product_id = s.product_id
            order by p.purchase_date desc
           ) a purchaseprice
from sales s

And the performance should be fine if you have an index on purchase(product_id, purchase_date) or even purchase(product_id, purchase_date, ave_cost_price).