home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.stat
- Path: sparky!uunet!pipex!warwick!pavo.csi.cam.ac.uk!camcus!nmm1
- From: nmm1@cus.cam.ac.uk (Nick Maclaren)
- Subject: Re: bug in fortran RNG
- Message-ID: <1992Nov6.172324.23826@infodev.cam.ac.uk>
- Sender: news@infodev.cam.ac.uk (USENET news)
- Nntp-Posting-Host: bootes.cus.cam.ac.uk
- Organization: U of Cambridge, England
- References: <1992Nov5.205359.29733@cbfsb.cb.att.com>
- Date: Fri, 6 Nov 1992 17:23:24 GMT
- Lines: 23
-
- In article <1992Nov5.205359.29733@cbfsb.cb.att.com>, rizzo@cbnewsf.cb.att.com (anthony.r.rizzo) writes:
- |> Some time ago, during a discussion of random number generators,
- |> at least one individual mentioned that the RNG in "Numerical Recipes,"
- |> perhaps the fortran version, had a bug in it. Does anyone recall
- |> what that bug was and how to correct it? Does anyone recall
- |> which of the various RNGs in "Numerical Recipes" has the bug?
- |> Thanks!
-
- I might have been the person, but I didn't say that there was a bug in any of
- the basic generators in Numerical Recipes. I said that most of them were
- ghastly, and fail several of the standard tests very badly. The only one
- that is any good is the DES-based one, and I suggest using more efficient
- code (such as Eric Young's). Alternatively, use a reliable traditional
- generator, such as Wichmann and Hill's.
-
-
- Nick Maclaren
- University of Cambridge Computer Laboratory,
- New Museums Site, Pembroke Street,
- Cambridge CB2 3QG, England.
- Email: nmm1@cus.cam.ac.uk
- Tel.: +44 223 334761
- Fax: +44 223 334679
-