Mailing-List: contact issues-help@commons.apache.org; run by ezmlm
Precedence: bulk
Reply-To: issues@commons.apache.org
Date: Mon, 27 Aug 2012 22:00:08 +1100 (NCT)
From: "Gilles (JIRA)" <jira@apache.org>
To: issues@commons.apache.org
Message-ID: <918862009.974.1346065208141.JavaMail.jiratomcat@arcas>
In-Reply-To: <1182901071.28807.1345433918089.JavaMail.jiratomcat@arcas>
Subject: [jira] [Commented] (MATH-849) Precision improvement in class Gamma
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/MATH-849?page=3Dcom.atlassian.j=
ira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D134423=
59#comment-13442359 ]=20

Gilles commented on MATH-849:
-----------------------------

Why can't you keep a single implementation (the one that works in both rang=
es)?
IOW, why can't the new implementation be used for issue MATH-753?
Is it maybe an issue of performance? By the way, did you benchmark the alte=
rnative implementations? If there is a "big" difference, maybe that some us=
ers will be willing to trade off accuracy for performance. That could be in=
dicated in the Javadoc.

               =20
> Precision improvement in class Gamma
> ------------------------------------
>
>                 Key: MATH-849
>                 URL: https://issues.apache.org/jira/browse/MATH-849
>             Project: Commons Math
>          Issue Type: Improvement
>            Reporter: S=C3=A9bastien Brisard
>            Assignee: S=C3=A9bastien Brisard
>            Priority: Minor
>              Labels: accuracy, special-functions
>             Fix For: 3.1
>
>         Attachments: MATH-849-Lanczos_vs_NSWC.zip, MATH-849.patch, MATH-8=
49.zip
>
>
> New unit tests (see below) show that (on the set of values tested) the cu=
rrent accuracy of {{Gamma.logGamma(double)}} is about 130 ulps.
> The [NSWC Library of Mathematical Functions|http://www.dtic.mil/docs/cita=
tions/ADA476840] proposes an implementation of this function with much high=
er accuracy: on the same set of values, the error goes down to about 3 ulps=
.
> This library is available [here|http://www.ualberta.ca/CNS/RESEARCH/Softw=
are/NumericalNSWC/site.html], and is license-free [see this thread|http://m=
arkmail.org/thread/fg4y4dfdnfty5hov]. It is proposed to port this implement=
ation in Commons-Math. This in turn could lead to a resolution of MATH-738.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrato=
rs
For more information on JIRA, see: http://www.atlassian.com/software/jira