Question

I am trying to write a stored procedure ( using SQL Server Management Studio 2008 R2) to retrieve the maximum measurement value from a table. This seemed like an easy thing to do, so I wrote a short stored procedure to get the MAX. However, when I ran the procedure, it returned a 0 rather than the expected 0.018 value.

When I hover my mouse over the MAX function, it displays “built-in function MAX(expression) RETURNS int”. So it seems to be casting my result as an int. This is rather annoying.

I guess I’ll have to calculate the maximum the long way unless anyone here has an easier solution. Any ideas?

ALTER PROCEDURE [dbo].[GetMaxOscillation] (@SerialNumber varchar (16), @Date datetime)
AS
BEGIN
    SET NOCOUNT ON;

    DECLARE @MaxOsc DECIMAL(18,6)

    SELECT @MaxOsc = MAX(Value)
    FROM LoadListingOscillations
    WHERE SerialNumber=@SerialNumber AND Date=@Date AND Value IS NOT NULL


    RETURN @MaxOsc

    SET NOCOUNT OFF
END

Answer 1

You have to do MAX(CAST(Value AS DECIMAL(18,6)))

I am guessing that the Value is not a DECIMAL, so you must cast it before the assignment

Here is the SQLFiddle to prove that if the table is set up as a Decimal then this will work

Here is a SQLFiddle that shows that the cast is needed

UPDATE

If Value is indeed a Decimal(18,6), then you need to make sure of your table values. Because, as you can see from my examples, this should work

The last thing that is probably more your problem is that you are returning an integer. I do not think you can return anything but an integer in SQL Server. You will have to do a SELECT @MaxOsc or use an OUTPUT variable