I spoke with my colleague who did most of the testing, and he has informed me that much of the hessian sensitivity actually came from a separate program (based on Numerical Recipes in C++ code) that did not use optim(), after having stopped using optim() due to speed issues.<div>
<br></div><div>In my experience with optim, the reltol argument has improved important in this regard. Very small changes in the parameter estimates at the converged solution (influenced by reltol) can lead to different standard error estimates by inverting the hessian, especially for parameter estimates close to zero (as vulnerability coefficients can be in many models with such a feature). It is a limitation of the finite difference method for computing the hessian based on optimal parameter estimates.</div>
<div><br></div><div><br></div><div><br></div><div>Chris</div><div><br></div><div><br clear="all"><br>-----------------------------<br>Chris Gast<br><a href="mailto:cmgast@gmail.com">cmgast@gmail.com</a><br>
<br><br><div class="gmail_quote">On Wed, Jun 16, 2010 at 11:05 PM, Rubén Roa <span dir="ltr"><<a href="mailto:rroa@azti.es">rroa@azti.es</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
________________________________<br>
<br>
De: <a href="mailto:users-bounces@admb-project.org">users-bounces@admb-project.org</a> [mailto:<a href="mailto:users-bounces@admb-project.org">users-bounces@admb-project.org</a>] En nombre de Chris Gast<br>
Enviado el: miércoles, 16 de junio de 2010 21:11<br>
Para: Arni Magnusson<br>
CC: <a href="mailto:r-help@r-project.org">r-help@r-project.org</a>; <a href="mailto:users@admb-project.org">users@admb-project.org</a><br>
Asunto: Re: [ADMB Users] an alternative to R for nonlinear stat models<br>
<div class="im"><br>
Hi Arni (and others),<br>
My dissertation work involves use (and extension) of models of the same ilk (sometimes exactly the same) as those described by Nancy Gove and John Skalski in their 2002 article. I began with R, and moved to my own home-brewed C/C++ programs for the sake of of speed when fitting models and real and simulated data. In addition, we found that the estimated standard errors (based on the inverse hessian output from optim()) were very sensitive to tolerance criteria--often changing orders of magnitude.<br>
<br>
<br>
</div>Hi,<br>
Regarding the last bit, optim() has several methods (Nelder-Mead, simulated annealing, conjugate gradient, etc). It is interesting to me which method produced what result with the standard errors from the inverse Hessian. Can you briefly ellaborate?<br>
Thanks<br>
Rubén<br>
<br>
____________________________________________________________________________________<br>
<br>
Dr. Rubén Roa-Ureta<br>
AZTI - Tecnalia / Marine Research Unit<br>
Txatxarramendi Ugartea z/g<br>
48395 Sukarrieta (Bizkaia)<br>
SPAIN<br>
<br>
</blockquote></div><br></div>