Re: ranlux (new version)

[James Theiler <jt@lanl.gov>]

Hi Martin,
  And thanks a lot for the code.  (Also for the comment about DIEHARD.)  
I'm still interested in RNG's, but haven't done too much with them lately.  
I'm off to a conference next week, so won't be able to look at your code
right away, but I've forwarded it to Brian Gough and Mark Galassi, the two
most likely to want to do something with it. If anybody else on
gsl-discuss would like the code, I'd be happy to fwd it along as well,
though you might prefer going through Martin, so he can see how well
appreciated his efforts are.

regards,
jt

On Thu, 15 Jul 1999, Martin Luescher wrote:

] 
] 
] Hello James (if I may),
] 
] I am not sure whether you are still working on the random number generator
] suite included in the GSL, but if so I thought you might be interested to
] hear that there exists a simple way to accelerate the ranlux generator by
] a factor 2 to 3 using double precision arithmetic. This could be an
] important improvement in practice since the generator is otherwise rather
] slow and maybe even too slow for some applications. 
] 
] In the attachment I include the relevant documentation and an ANSI C
] program with the new code. There is now also an option to output double
] precision random numbers (48 random bits) although I doubt that random
] numbers of this quality will ever be required in field theory simulations
] or Monte Carlo integration routines. The program is completely safe and
] fully tested.
] 
] As far as statistical tests are concerned I would like to add that the
] "minimum distance" program in the DIEHARD test suite (as obtained from
] Marsaglia's website) contains a bug and yields incorrect results. 
] 
] Let me know if I can be of any help with this.
] 
] Martin
] 
] ********************************************************************************
] Martin Luescher                 phone:  +49 40 8998 2412
] DESY (Theory Group)             fax:    +49 40 8998 2777
] Notkestrasse 85                 e-mail: luscher@mail.desy.de
] D-22603 Hamburg 
] ********************************************************************************
] 
] 

---------------------------------------------
James Theiler                     jt@lanl.gov
MS-D436, NIS-2, LANL        tel: 505/665-5682
Los Alamos, NM 87545        fax: 505/665-4414
----- Space and Remote Sensing Sciences -----