Objet : Help-mcsim Digest, Vol 5, Issue 1

Hi !

We used a loguniform on sigma (the SD), and you can do the same with MCSim.

Frederic Bois

>>> <help-mcsim-request <at> gnu.org> Jeudi 29 mai 2003 18:04:41 >>>
Today's Topics:

1.  (Eric Hack)

In the Gelman, Bois, and Jiang 1996 JASA article entitled Physiological
Pharmacokinetic Analysis Using Population Modeling and Informative Prior
Distributions, the measurement error model uses the standard noninformative
prior distribution for the variances, p(sigma1^2, sigma2^2) proportional to
sigma1^(-2) sigma2^(-2).  How do I implement this in MCSIM?

Thanks in advance,

Eric