On Sun, Aug 3, 2008 at 7:20 PM, Dan Hatton <vi5u0-maxima at yahoo.co.uk> wrote:
>>>> I may try doing a MCMC sampling method for getting a bayesian
>>>> posterior sample from the coefficients. Has anyone done anything
>>>> like that in maxima before?
>
> I seem to have missed the start of this conversation, so this may not
> be quite what you're looking for (in particular, it wasn't done in
> Maxima!), but...
>
> The archive
> <http://www.bib.hatton.btinternet.co.uk/dan/Natural_Sciences/PER_CoCu.tar.gz>;
> contains (among other things) a GPL'd Perl script, PER_CoCu/main,
> which has subroutines for using the Metropolis-Hastings algorithm,
> with leapfrog proposal density, to obtain posterior expectations and
> standard deviations of model parameters, and a marginal likelihood for
> comparing the model with other models. I'm afraid it's a bit untidy
> and not very user-friendly, but you might be able to do something with
> it.
>
> Some explanation of what I was up to in this script, and references to
> the literature, can be found in the section of my thesis at
> <http://www.bib.hatton.btinternet.co.uk/dan/Natural_Sciences/PER_CoCu/node65.html>;.
> Literature citations in shorthand in comments of the script are also
> expanded in the bibliography of that thesis.
R is more suitable for Bayesian estimation. See:
http://www.r-project.org/
Paul