This is the mail archive of the gsl-discuss@sourceware.org mailing list for the GSL project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: Multiprecision arithmetic

From: Patrick Alken <alken at colorado dot edu>
To: <gsl-discuss at sourceware dot org>
Date: Tue, 24 Nov 2015 09:00:33 -0700
Subject: Re: Multiprecision arithmetic
Authentication-results: sourceware.org; auth=none
Authentication-results: spf=none (sender IP is ) smtp dot mailfrom=patrick dot alken at colorado dot edu;
References: <BLUPR16MB0196C8B54FECCB5A91C42236C9060 at BLUPR16MB0196 dot namprd16 dot prod dot outlook dot com>
Spamdiagnosticmetadata: NSPM
Spamdiagnosticoutput: 1:23

Yes for now I think it would add too much complexity to GSL, so weprefer to use double precision.

Thanks for your work on the recent bugs - I haven't had a chance to lookat the patches but I'm hoping to get to this soon.


Patrick

On 11/24/2015 07:18 AM, Bill Maier wrote:

I notice than in the write-up for bug #45746 it is suggested that multiprecision
arithmetic be used in a portion of the calculation to produce a more precise
result. Is this something that would be considered in GSL? I did a search
through the code for mpfr or gmp and did not find either. In fact I couldn't find
any code where multiprecision arithmetic is currently being used. Is it GSL
policy to keep all internal calculations in double precision?

-Bill Maier

References:
- Multiprecision arithmetic
  - From: Bill Maier

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]