RE: SQL ROUND function on a FLOAT

Why do you use the float? I always thought that this is an approximate
value. 

-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx
[mailto:rpg400-l-bounces@xxxxxxxxxxxx] On Behalf Of Mark Adkins
Sent: Thursday, August 24, 2006 11:34 AM
To: rpg400-l@xxxxxxxxxxxx
Subject: SQL ROUND function on a FLOAT

This isn't so much a question but more like a "what the?!".

If I have the float value "3.4173900000000001E+000" and issue a "ROUND( 
3.4173900000000001E+000, 2 )", I get (drum roll)
3.4199999999999999E+000. 
Cast as a dec with 2 decimal places and viola!  You get 3.41739 rounded
to 
3.41.

This statement shows it all. 

SELECT FLOAT(3.41739) AS NUM, 
       ROUND( FLOAT(3.41739), 2 ) AS ROUNDED_NUM, 
       CAST( ROUND( FLOAT(3.41739), 2 ) AS DEC(5,2) )
        AS CASTED_ROUNDED_NUM 
FROM SYSIBM/SYSDUMMY1 

Output:
                    NUM               ROUNDED_NUM  CASTED_ROUNDED_NUM
3.4173900000000001E+000   3.4199999999999999E+000          3.41 

Of course I know to cast it to 3 decimal places before the round, but 
still...

Mark