Mailing-List: contact issues-help@commons.apache.org; run by ezmlm
Precedence: bulk
Reply-To: issues@commons.apache.org
Date: Tue, 7 Jun 2011 08:47:58 +0000 (UTC)
From: "Mikkel Meyer Andersen (JIRA)" <jira@apache.org>
To: issues@commons.apache.org
Message-ID: 
 <1437631674.3124.1307436478967.JavaMail.tomcat@hel.zones.apache.org>
In-Reply-To: 
 <1440803169.68751.1307212727362.JavaMail.tomcat@hel.zones.apache.org>
Subject: [jira] [Commented] (MATH-585) Very slow generation of gamma random
 variates
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/MATH-585?page=3Dcom.atlassian.j=
ira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D130453=
26#comment-13045326 ]=20

Mikkel Meyer Andersen commented on MATH-585:
--------------------------------------------

I know that it is indeed important, and I will look into it as soon as time=
 allows me to. R is GPL, so using their implementation is a no-go. But the =
help file provides the methods they have implemented:
{quote}
=E2=80=98rgamma=E2=80=99 for =E2=80=98shape >=3D 1=E2=80=99 uses

Ahrens, J. H. and Dieter, U. (1982).  Generating gamma variates by
a modified rejection technique.  _Communications of the ACM_,
*25*, 47-54,

and for =E2=80=980 < shape < 1=E2=80=99 uses

Ahrens, J. H. and Dieter, U. (1974).  Computer methods for
sampling from gamma, beta, Poisson and binomial distributions.
_Computing_, *12*, 223-246.
{quote}

Again, thanks for reporting this - we'll do our best to improve our impleme=
ntation.

> Very slow generation of gamma random variates
> ---------------------------------------------
>
>                 Key: MATH-585
>                 URL: https://issues.apache.org/jira/browse/MATH-585
>             Project: Commons Math
>          Issue Type: Improvement
>    Affects Versions: 2.2, 3.0
>         Environment: All
>            Reporter: Darren Wilkinson
>              Labels: Gamma, Random
>   Original Estimate: 6h
>  Remaining Estimate: 6h
>
> The current implementation of gamma random variate generation works, but =
uses an inversion method. This is well-known to be a bad idea. Usually a ca=
refully constructed rejection procedure is used. To give an idea of the mag=
nitude of the problem, the Gamma variate generation in Parallel COLT is rou=
ghly 50 times faster than in Commons Math.=20

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira