Question

I am making a storedprocedure which has a SELECT clause selecting multiple colums from some joined tables. For one of these colums the value is depending on a condition with CASE.

SELECT DISTINCT
 table1.col1 as var1,
 table2.col2 as var2,
 CASE WHEN (FLOOR(table3.col3/0.2)*0.2) > 20 THEN 20
      WHEN (FLOOR(table3.col3/0.2)*0.2) <= 20 THEN FLOOR(table3.col3/0.2)*0.2
      ELSE table3.col4 -- selecting another value is the 
                       -- value in table3.col3 is null
      END as var3 
FROM ... 
WHERE ...

As you can see in the select I’m doing three times the FLOOR(table3.col3/0.2)*0.2 calculation. Is it possible to store this calculated value in a variable WITHIN THE SELECT and then use this variable in the conditions?

Thanks in advance,

Jeroen

Answer 1

No, you can reduce it to 2 calls like below. But whatever else you will do will be more costly then actually doing it like this, as what you do there is very fast. You can think on something to improve, when you actually have some heavy operations. This is just nothing. You might also make a get_min function but again I do not think it will be worth it.

With 2 calls:

case when table3.col3 then table3.col4
     when (FLOOR(table3.col3/0.2)*0.2) > 20 THEN 20
     else FLOOR(table3.col3/0.2)*0.2
end

With min function:

case when table3.col3 then table3.col4
     else dbo.get_min(FLOOR(table3.col3/0.2)*0.2,20)
end

As a side note, the min function I am referring to has to be built (user defined function).