From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 1308 invoked by alias); 7 Jan 2011 20:49:46 -0000 Received: (qmail 1294 invoked by uid 22791); 7 Jan 2011 20:49:45 -0000 X-SWARE-Spam-Status: No, hits=-2.2 required=5.0 tests=AWL,BAYES_00,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from mail.phy.duke.edu (HELO mail.phy.duke.edu) (152.3.182.2) by sourceware.org (qpsmtpd/0.43rc1) with SMTP; Fri, 07 Jan 2011 20:49:32 +0000 Received: from localhost (localhost [127.0.0.1]) by mail.phy.duke.edu (Postfix) with ESMTP id 9D900780EC; Fri, 7 Jan 2011 15:47:35 -0500 (EST) Received: from mail.phy.duke.edu ([127.0.0.1]) by localhost (mail.phy.duke.edu [127.0.0.1]) (amavisd-new, port 10026) with LMTP id K3rt-C3UrVYd; Fri, 7 Jan 2011 15:47:35 -0500 (EST) Received: from lilith (unknown [10.180.136.65]) by mail.phy.duke.edu (Postfix) with ESMTP id 886B5780EA; Fri, 7 Jan 2011 15:47:35 -0500 (EST) Date: Fri, 07 Jan 2011 20:49:00 -0000 From: "Robert G. Brown" To: Brian Gough cc: GSL Discussion list Subject: Re: Three random number generators... In-Reply-To: <43r5cojy1y.wl%bjg@gnu.org> Message-ID: References: <43r5cojy1y.wl%bjg@gnu.org> User-Agent: Alpine 2.02 (LFD 1266 2009-07-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed 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-q1/txt/msg00004.txt.bz2 On Fri, 7 Jan 2011, Brian Gough wrote: > At Fri, 7 Jan 2011 13:47:41 -0500 (EST), > Robert G. Brown wrote: >> >> I have three random number generators that I think ought to go into the >> GSL. One is AES -- which is pretty self explanatory -- fairly fast, >> cryptographic strength. >> ... >> Basically, I'm suggesting that it is time to update the random number >> generators in the GSL to make them more consistent with state of the art >> in speed and tested randomness. > > Thanks for that, I am happy to add the AES and KISS (or an enhanced > version of it). Instead of seeding with mt19937, I think it would be > preferable to generate the seeds inline and have the generator > standalone if possible. Maybe the 69069 LCG used elsewhere would > be sufficient(?), otherwise a strong hash function. Either one would be fine with me (it was just so EASY to use mt because the GSL is so EASY to use:-). I'll get a recent copy of the sources and see what you use IN the mt generators, which also require some work to initialize IIRC. That might take me a few days. I also need to do some careful testing...;-) I'll send them on when done. rgb Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb@phy.duke.edu