Date:         Sat, 4 Dec 1999 12:10:31 +1100
Reply-To:     Tim Churches <tchur@BIGPOND.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Tim Churches <tchur@BIGPOND.COM>
Subject:      Re: Calculation Precision Limits
Comments: To: David Cassell <cassell@MERCURY.COR.EPA.GOV>
Content-Type: text/plain; charset=us-ascii

David Cassell wrote:
> And Math::BigFloat can be made even easier for the SAS user,
> since all the basic math operations are overloaded if you
> just declare your big floats like this:
>     $float = new Math::BigFloat "2.123123123123123123123123123123123";

David,

Thanks for pointing this out. My original example in Perl can in fact be
written thus:

use Math::BigFloat;
$ten = new Math::BigFloat "10";
$float = $ten / 3 ;
print "10 / 3 = $float\n" ;
$tenagain = $float * 3 ;
print "$float * 3 = $tenagain\n" ;

For Windows 9x/NT/2k, a very convenient Perl distribution which includes
Math::BigFloat and which comes with a nice installation routine can be
obtained from http://www.activestate.com

Most Linux distributions include Perl, and Perl is available in source
or binary form for all Unix flavours, VAX/VMS, BeOS, Plan9 etc etc. I
don't think that there is a Perl distribution for IBM mainframes, but
apparently Linux is being ported to IBM mainframes - it is alreday
running in a VM (virtual machine) but is also being ported to run on the
bare metal of the big iron. So we will soon have Linux for IBM
mainframes but still no SAS for Linux...

Tim Churches