This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: rand_rate
- From: Linas Vepstas <linas at austin dot ibm dot com>
- To: "Robert G. Brown" <rgb at phy dot duke dot edu>
- Cc: GSL Discussion list <gsl-discuss at sources dot redhat dot com>
- Date: Wed, 10 Nov 2004 12:58:02 -0600
- Subject: Re: rand_rate
- References: <Pine.LNX.4.58.0411092336350.12471@lilith.rgb.private.net>
On Tue, Nov 09, 2004 at 11:50:12PM -0500, Robert G. Brown was heard to remark:
> I mentioned my random number tester program,
>
> http://www.phy.duke.edu/~rgb/General/rand_rate.php
>
> outputting a "report" that includes a p-value (or set of p-values) and a
> fairly arbitrary conclusion about whether or not the rng is likely to be
> questionable for people who can't interpret p-values, for the test being
> run.
>
> It is still very much a project in progress -- I'm missing lots (most)
> of the STS and diehard tests -- but it does provide a framework for
So what does it test for currently? (could you update the web page to
state this?) I've been fiddling with something that seems to be
like a prng (it seems to pass basic tests) but its not a traditional
polynomial curve/cryptographic prng; I want to understand if/how it
falls down. (its rather bizarre) I know how its weak in certain cases;
but want to understand if/how it might pass traditional tests.
--linas