Link to home
Start Free TrialLog in
Avatar of dabestprogrammerindauniverse
dabestprogrammerindauniverse

asked on

Banker's Rounding

MS SQL Server has the Round function that rounds-off of mid-values (0.5, 5, 50, 500, etc.) up. Is there a function that will round-off mid-values according to Banker's Rounding, which will round of mid-values always to the even value? Anyone has one?
0.5->0
1.5->2
2.5->2
3.5->4
Thanks! :)
ASKER CERTIFIED SOLUTION
Avatar of tomcat75
tomcat75

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of tomcat75
tomcat75

maybe an explanation will help

if
VAR-ROUND(VAR,0)=-0.5 (for 0.5 fractions)
if
FLOOR(VAR)/2=FLOOR(FLOOR(VAR)/2) --the number is even
then
FLOOR(VAR),
else
ROUND(VAR,0,
endif
ROUND(VAR,0)
endif


SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of dabestprogrammerindauniverse

ASKER

i will try your solutions out and will then award my points :)
thanks to everyone.
Hi dabestprogrammerindauniverse,

This function will round up or down to the degree you specify.

-- Round up to the the nearest cent
-- Set @Somevarmoney = dbo.RoundToNearest(@SomeVarmoney,.01,1)

-- Round down to the the nearest cent
-- Set @Somevarmoney = dbo.RoundToNearest(@SomeVarmoney,.01,0)


CREATE Function [dbo].[RoundToNearest] (@Amt decimal(38, 19), @RoundAmt decimal(38, 19), @Direction bit)
-- @Direction=1 for RoundUp, @Direction=0 for RoundDown
RETURNS decimal(38, 19) AS  
BEGIN
  RETURN (CASE WHEN CONVERT(BIGINT, @Amt/@RoundAmt)=@Amt/@RoundAmt THEN @Amt ELSE (CONVERT(BIGINT, @Amt/@RoundAmt)+@Direction)*@RoundAmt END)
END


Alan
May I ask the reason for this recommendation.

Not discounting alanwarren's solution, which appears in essence very similar to the three previously posted solutions albeit posted 25 days after those solutions, and indeed 11 days after the asker posted they would try the already proffered solutions.

I believe the point should be shared between Tomcat75, xabimond & myself as these were posted prior to the askers final post and indeed in good hast to the askers question.

I look forward to your reply.