Bill
I keep meaning to get into this field properly, meanwhile using
BUGS on occasion - _very_ occasionally in fact. You probably
know that you can link BUGS and R with RBugs and then J and R -
don't know if it's practical to explore that line.
I expect he'll reply to your forum posting in his own right, but
it might be worth dropping Ewart Shaw a direct line. As far as
I know he's still doing Bayesian & MCMC stuff at Warwick University.
He's at [EMAIL PROTECTED]
Good Luck
Mike
Bill Harris wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Just curious: is there a public MCMC (Markov Chain Monte Carlo) engine
in J? I'm using MCSim a bit and starting to explore OpenBugs, but
seeing
http://www.stat.columbia.edu/~cook/movabletype/archives/2007/08/hierarchical_ba.html
this morning made me wonder again about J. The comment about "scaling
to large quantities of data are vital" seems to be a good fit for J.
If there is something (or if it turns out to be a 3-liner), it would be
interesting to know about it here, and it might be interesting to add as
a comment to that blog entry.
Thanks,
Bill
- --
Bill Harris http://facilitatedsystems.com/weblog/
Facilitated Systems Everett, WA 98208 USA
http://facilitatedsystems.com/ phone: +1 425 337-5541
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (MingW32)
Comment: For more information, see http://www.gnupg.org
iD8DBQFG1uNF3J3HaQTDvd8RApBhAKCBUTyD5l5Xrq730GSoKiLetGyrYgCeMkjP
vkGMvYGRbaf/VTL4KSJnl/Y=
=LzcT
-----END PGP SIGNATURE-----
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
