Date: Wed, 16 Jan 2002 09:51:13 -0000
Reply-To: Nigel.Pain@SCOTLAND.GSI.GOV.UK
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Nigel Pain <Nigel.Pain@SCOTLAND.GSI.GOV.UK>
Subject: Re: mod-function
Content-Type: multipart/alternative;
*****************************************************************************
This email and any files transmitted with it are intended solely
for the use of the individual or entity to whom they are addressed.
*****************************************************************************
> -----Original Message-----
> From: John Whittington [mailto:John.W@MEDISCIENCE.CO.UK]
> Sent: 15 January 2002 15:17
> Subject: Re: mod-function
>
>
> At 08:29 15/01/02 -0500, Roger Lustig wrote (in part):
>
> >Engelbert:
> >I get the expected zeroes under Win2K/8.2. *Very* strange!
> >What aspects of your program have you varied? Does the LENGTH
> >statement change things, for instance?
> >Needless to say, mod(anything,n) should *never* return a value of n.
>
> That is indeed correct; mod(anything,n) can never return a
> value equal to
> or greater than n. However, I'm sure the point with
> Englebert's example is
> that what the mod() function is returning only _appears_ to
> be 'n' because
> of the precision with which it is being displayed. As we
> discuss so often
> here, 25.4 cannot be stored precisely in binary, so 25.4*50
> will not be
> EXACTLY 1270 within the machine.
And Roger is correct to question the length statement. This gives all
numeric variables a length of 4 bytes, which considerably reduces the
precision with which decimals can be stored. It is important to use the
maximum length (8) for numbers which may have a fractional portion to them
so that precision is maximised.
***************************************************
Nigel Pain
Scottish Executive Development Department
Business Support Unit
Victoria Quay
EDINBURGH
EH6 6QQ
UK
Tel +44 131 244 7237
Fax +44 7092 014235
Mailto:nigel.pain@scotland.gsi.gov.uk
Website: http:\\www.scotland.gov.uk
[text/html]
