From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 15030 invoked by alias); 6 Jun 2011 02:20:26 -0000 Received: (qmail 15019 invoked by uid 22791); 6 Jun 2011 02:20:25 -0000 X-SWARE-Spam-Status: No, hits=2.3 required=5.0 tests=BAYES_50,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_ENVFROM_END_DIGIT,FREEMAIL_FROM,RCVD_IN_DNSWL_NONE,RFC_ABUSE_POST,T_RP_MATCHES_RCVD,T_TO_NO_BRKTS_FREEMAIL X-Spam-Check-By: sourceware.org Received: from nm27.bullet.mail.ne1.yahoo.com (HELO nm27.bullet.mail.ne1.yahoo.com) (98.138.90.90) by sourceware.org (qpsmtpd/0.43rc1) with SMTP; Mon, 06 Jun 2011 02:20:10 +0000 Received: from [98.138.90.56] by nm27.bullet.mail.ne1.yahoo.com with NNFMP; 06 Jun 2011 02:20:09 -0000 Received: from [98.138.88.233] by tm9.bullet.mail.ne1.yahoo.com with NNFMP; 06 Jun 2011 02:20:09 -0000 Received: from [127.0.0.1] by omp1033.mail.ne1.yahoo.com with NNFMP; 06 Jun 2011 02:19:53 -0000 Received: (qmail 16523 invoked by uid 60001); 6 Jun 2011 02:19:53 -0000 Message-ID: <561449.86378.qm@web110509.mail.gq1.yahoo.com> Received: from [24.44.10.80] by web110509.mail.gq1.yahoo.com via HTTP; Sun, 05 Jun 2011 19:19:53 PDT Date: Mon, 06 Jun 2011 02:20:00 -0000 From: Z F Subject: Re: accuracy of gsl_cdf_binomial_P To: gsl-discuss@sourceware.org, well@wheatstone-analytics.com In-Reply-To: <4DEB7EA2.7090500@superlink.net> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Mailing-List: contact gsl-discuss-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gsl-discuss-owner@sourceware.org X-SW-Source: 2011-q2/txt/msg00018.txt.bz2 Dear Well Howell, --- On Sun, 6/5/11, Well Howell wrote: > An interesting (but "homework-like" > ~;) question - and fun to answer too. >=20 > Anyway, I'd probably compare GSL results with those from > other sources. >=20 > I had easy access to gsl_cdf_binomial_P (v 1.14),=A0 R > pbinom(k,n,p),=20 > binomCDF > (Excel 2007) and dcdflib (Fortran - Brown, Lovato & > Russel; U. Texas;=20 > November, 1997). >=20 > For a sample size of n=3D1000, a trial probability of p=3D0.01 > and number of=20 > successes of > s=3D1 thru 40, the CDF values from dcdclib and the R 2.13.0 > stats package=20 > pbinom() > function (http://cran.r-project.org/) show no > difference. >=20 Thank you very much for your reply.=20 It seems I was not clear with my question. I am not looking for a comparison with other libraries, but rather for information regarding the approximations used to obtain the values of CDF. What I am afraid of is that a Gaussian approximation is used for a large sample, rendering values in the tails of the distribution error-prone. I someone could provide any info on the subject or maybe point in the "righ= t direction" , I would highly appreciate it. Thanks again ZF >=20 > On 6/2/2011 12:49 AM, Z F wrote: > > Hello everybody, > > > > I was wondering if someone could comment on the > accuracy of gsl_cdf_binomial_P() function gsl implementation > for large n (n is about a few thousand). > > for different values of p and when the result of cdf > is in the tails ( small less then 0.05 and large -- above > 0.95) > > > > Thank you very much > > > > ZF > > > > >=20 >