Clever Geek Handbook

Getting rid of negative null value for float

With SQL Server 2008R2, I get a sign difference when rounding "the same" number as decimal or float.

Sample code showing the difference:

    DECLARE @f float = -0.00001;
    DECLARE @d decimal = -0.00001;
    PRINT CONVERT(varchar,ROUND(@f, 4))  --- outputs -0
    PRINT CONVERT(varchar,ROUND(@d, 4))  --- outputs 0

I need to round the float value to 0, not -0.

I think it boils down to the IEEE 754 specification, which defines two zero representations (positive and negative) whereas decimal numbers only have one zero. I think the float equality operator does the job, but I have to compare two AS floating point strings for compatibility reasons.

Is there a / function to get rid of negative zeros (at least when converting to a string)?

+3


source to share


2 answers


Adding 0.0 to the value fixed the problem. Thanks tmyblebu for the answer.



     PRINT CONVERT(varchar,ROUND(@f + 0.0, 4))
+2


source


@Pascal Try below code. Let me know if it helped you.



DECLARE @f float = -0.00001;
    DECLARE @d decimal = -0.00001;
    PRINT abs(ROUND(@f, 4))  --- outputs 0
    PRINT abs(ROUND(@f, 4))   --- outputs 0
-1


source






More articles:


All Articles