Dear Andrew and R-list,
I guess Fournier is addressing the properties of the numerical routines
underlying the various packages, not the statistical properties of the MLE
itself.
For this purpose using a small tricky dataset makes sense. Clearly,
a true unique MLE exists (except in pathological
Dear Hans,
these are interesting points. I guess that I'm approaching it from
the point of view of a decision: I'd be more comfortable using a
fitting routine that has stability under a wide range of identifiable
circumstances. Obtaining the MLE exactly in any instance is a function
of the data
Do Users of Nonlinear Mixed Effects Models Know
Whether Their Software Really Works?
Lesaffre et. al. (Appl. Statist. (2001) 50, Part3, pp 325-335)
analyzed
some simple clinical trials data using a logistic random effects
model. Several packages and
On Thu, 13 Oct 2005, dave fournier wrote:
Internal Virus Database is out-of-date.
Talk about not being careful!
--
Brian D. Ripley, [EMAIL PROTECTED]
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861
Dave,
that's an interesting start for a comparison. Let me point out some
ways that you might construct a compelling argument. Of course, these
aren't exhaustive, and others may well provide further depth.
1) If I understand correctly, you're trying to estimate parameters
from a real