RE: SQL RAND function question

[Vernon Hamberg <vhamberg@xxxxxxxxxxx>]

Hi Charles

The word "granularity" comes to mind - if multiplying and rounding, 
the evenness might be compromised. But I would still expect each 
bucket to have about the same number or elements, given enough iterations.

I've not been completely happy with the SQL RAND() function, but 
maybe I did something to affect it. All these things ARE 
pseudo-random, after all, and the algorithm can make a big 
difference. But now I speak with less than exhaustive experience.  ;-)

Regards
Vern

At 07:19 AM 2/13/2008, you wrote:


> Probably should have said, "by definition random numbers should be 
> evenly distributed".
>
> The RAND() function does AFAIK provide an even distribution from 
> [0,1].  The problem is what happens
> when you multiply that by N.
>
>
> Charles Wilt
> Software Engineer
> CINTAS Corporation - IT 92B
> 513.701.1307
>
> wiltc@xxxxxxxxxx
>
>
> > -----Original Message-----
> > From: midrange-l-bounces@xxxxxxxxxxxx [mailto:midrange-l-
> > bounces@xxxxxxxxxxxx] On Behalf Of Wilt, Charles
> > Sent: Wednesday, February 13, 2008 8:15 AM
> > To: Midrange Systems Technical Discussion
> > Subject: RE: SQL RAND function question
> >
> > Loyd,
> >
> > Yes it does.  BUT if you call it 1000's of times, it doesn't given you an
> > even distribution of numbers
> > between 0-11.
> >
> > By definition, RAND() should given an even distribution.
> >
> > Charles Wilt
> > Software Engineer
> > CINTAS Corporation - IT 92B
> > 513.701.1307
> >
> > wiltc@xxxxxxxxxx
> >
> >