Question

SQL 2005 :

I am working on set of SQL queries which generate a formula.

This Mathematical calculation, Example : (3000 -30)/1477*100 = 201.1 % has to be generated from set of Tables.

Value : 3000 is generated from this query ..
select sum(DaysSupply) as [Total Days Supply] from vOeOrderWide where patcustid = 4797 
Output is [Total Days Supply]


Value : 30 is generated from this query ..
select top 1 DaysSupply from VoeOrderwide where patcustid = 4797 order by datefilled desc
Output is [DaysSupply]

Value : 1477 is generated from these set of queries ...

declare @d1  datetime;
declare @d2  datetime;
set @d1= (select top 1 DateFilled from vOeOrderWide where patcustid = 4797 
order by DateFilled asc)
set @d2= (select top 1 DateFilled from vOeOrderWide where patcustid = 4797 order by DateFilled DESC )
select DATEDIFF(d,@d1,@d2) as [Days Between] 

Output is [Days Between]

I want to combine all these queries and generate the formula as ..

[Total Days Supply] - [DaysSupply]  /  [Days Between]  * 100

Answer 1

Hmm, I think I’d try something more along these lines:

WITH supply_data (Total_Days_Supply, Earliest_Date_Filled, Latest_Date_Filled) as (
                  SELECT SUM(DaysSupply), MIN(DateFilled), MAX(DateFilled)
                  FROM voeOrderWide
                  WHERE patCustId = 4797)
SELECT 100.00 * ((Total_Days_Supply - (SELECT TOP 1 DaysSupply
                                       FROM voeOrderWide
                                       WHERE patCustId = 4797
                                       AND DateFilled = LatestDateFilled
                                       ORDER BY DateFilled DESC)) 
                 / DateDiff(d, Earliest_Date_Filled, Latest_Date_Filled))
FROM supply_data

Please note that I do not have an SQL server instance to test this against.