<p>Russel and Anders,</p><p></p><p>I have not noticed this before, and unfortunately I can not tell why it happens. </p><p></p><p>Hans</p><br><div class="gmail_quote">On Mon, Mar 16, 2009 at 9:27 AM, Anders Nielsen <span dir="ltr"><<a href="mailto:anders@nielsensweb.org">anders@nielsensweb.org</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Dear Russel,<br>
<br>
I think that only Dave or Hans is able to tell you why exactly<br>
this happens, but I noticed something slightly related at one<br>
point, which is that the 'max gradient component' in the par<br>
file did not match the one in the screen output. In that case<br>
it helped if I used the -noinit flag when I ran the program.<br>
<br>
Cheers,<br>
<font color="#888888"><br>
Anders.<br>
</font><div><div class="h5"><br>
<br>
<br>
On Mon, 2009-03-16 at 19:48 +1300, <a href="mailto:r.millar@auckland.ac.nz">r.millar@auckland.ac.nz</a> wrote:<br>
> Dear ADMB'ers,<br>
><br>
> First, I'd like to give my thanks to the ADMB foundation team for making<br>
> ADMB freeware. Also, I'm really appreciative to Jon Schnute for the<br>
> PBSadmb package and the excellent installation instructions. I gave them<br>
> to a grad student a few days ago, and he was able to get ADMB up and<br>
> running without any further input from me (a first!).<br>
><br>
> In fact, I've been enjoying ADMB-RE so much that I might have come across<br>
> a little glitch. It is that the .std and .rep file will return slightly<br>
> different values for the std error of a random effect vector. For example,<br>
> in the orange tree example (obtained from ADMB webpages), I added a few<br>
> lines to the REPORT section. The .std file looks like<br>
><br>
> index name value std dev<br>
> 1 beta 5.3157e-02 1.5658e+01<br>
> 2 beta 1.9063e+00 3.5249e+01<br>
> 3 beta -7.9269e+00 2.7080e+01<br>
> 4 log_sigma 2.0596e+00 1.2910e-01<br>
> 5 log_sigma_u 3.4561e+00 3.2412e-01<br>
> 6 u -2.9564e+01 1.4740e+01<br>
> 7 u 3.1730e+01 1.4742e+01<br>
> 8 u -3.7196e+01 1.4759e+01<br>
> 9 u 4.0227e+01 1.4768e+01<br>
> 10 u -5.1974e+00 1.4700e+01<br>
><br>
> and the .rep file produces (with my obvious addition of code)<br>
><br>
> log_sigma = 2.05962<br>
> log_sigma_u = 3.45462<br>
> sigma = 7.84301<br>
> sigma_u = 31.6463<br>
><br>
> Note that log_sigma_u's are different between the two files. The<br>
> difference is not much here, but it is clearly not round-off error. I had<br>
> earlier been running a linear normal one-way RE model and the relative<br>
> difference was much greater. I also had the MLE from R function lmer, and<br>
> can say that unfortunately it is the .std file that is in error (as is the<br>
> .cor file). Also, I've found that changing the bounds on the parameter<br>
> specification (but keeping them inactive) of log_sigma_u altered its value<br>
> in the .std file, but not the .rep file.<br>
><br>
> Any ideas?<br>
><br>
> Regards,<br>
><br>
> Russell<br>
><br>
> _______________________________________________<br>
> Users mailing list<br>
> <a href="mailto:Users@admb-project.org">Users@admb-project.org</a><br>
> <a href="http://lists.admb-project.org/mailman/listinfo/users" target="_blank">http://lists.admb-project.org/mailman/listinfo/users</a><br>
<br>
_______________________________________________<br>
Users mailing list<br>
<a href="mailto:Users@admb-project.org">Users@admb-project.org</a><br>
<a href="http://lists.admb-project.org/mailman/listinfo/users" target="_blank">http://lists.admb-project.org/mailman/listinfo/users</a><br>
</div></div></blockquote></div><br>