[ADMB Users] Problems creating *.psv file in mcsave step? (Windows, ADMB versions 9 and 10)

[Larry Jacobson]

1) I am having a strange problem with the -mceval step and the psv file 
in particular.  I am running ADMB versions 9 and 10 (Borland) in Windows 
and Version 10 with the IDE.  The problem occurs in all of these 
environments.

To demonstrate this, I added one sdreport_number to the simple example 
file distributed with ADMB (tpl and dat files attached).

2) The -mcmc step runs fine and generates the hst file.  I start from a 
clean directory where I know I have full permissions to 
read/write/create files.
> simple -mcmc 1000
>
> Initial statistics: 2 variables; iteration 0; function evaluation 0
> Function value   3.6493579e+01; maximum gradient component mag  
> -3.6127e+00
> Var   Value    Gradient   |Var   Value    Gradient   |Var   Value    
> Gradient
>   1  0.00000 -3.61269e+00 |  2  0.00000 -7.27814e-01 |
>
>  - final statistics:
> 2 variables; iteration 7; function evaluation 19
> Function value   1.4964e+01; maximum gradient component mag  -7.0014e-05
> Exit code = 1;  converg criter   1.0000e-04
> Var   Value    Gradient   |Var   Value    Gradient   |Var   Value    
> Gradient
>   1  1.90909 -7.00140e-05 |  2  4.07818 -2.08982e-05 |
> Estimating row 1 out of 2 for hessian
> Estimating row 2 out of 2 for hessian
>  0.00944716 0.510258
> Initial seed value -36519
> -14.9642 -14.9642
>  mcmc sim 1  acceptance rate 0 0
> -15.7248 -15.7248
>  mcmc sim 201  acceptance rate 0.522388 0.525
> increasing step size -1.32761
> -15.486 -15.486
>  mcmc sim 401  acceptance rate 0.458853 0.395
> -15.0452 -15.0452
>  mcmc sim 601  acceptance rate 0.464226 0.475
> increasing step size -0.962971
> -22.0249 -22.0249
>  mcmc sim 801  acceptance rate 0.439451 0.365
> 0.428
> In write empirical covariance matrix
>  0.0314052 1.20311

> dir
> 12/16/2011  08:57 AM                56 admodel.cov
> 12/16/2011  08:57 AM                39 admodel.dep
> 12/16/2011  08:57 AM                56 admodel.hes
> 12/16/2011  08:57 AM                 0 eigv.rpt
> 12/16/2011  08:57 AM                 0 fmin.log
> 12/16/2011  08:57 AM                16 simple.bar
> 12/16/2011  08:57 AM               283 simple.cor
> 12/16/2011  08:56 AM             1,946 simple.cpp
> 09/30/2011  02:42 PM               172 simple.dat
> 12/16/2011  08:57 AM                32 simple.ecm
> 12/16/2011  08:57 AM                32 simple.eva
> 12/16/2011  08:56 AM           528,384 simple.exe
> 12/16/2011  08:57 AM               957 simple.hst
> 12/16/2011  08:56 AM             1,350 simple.htp
> 12/16/2011  08:57 AM               123 simple.log
> 12/16/2011  08:57 AM                 0 simple.mc2
> 12/16/2011  08:57 AM             8,000 simple.mcm
> 12/16/2011  08:56 AM           286,190 simple.obj
> 12/16/2011  08:57 AM               148 simple.par
> 12/16/2011  08:57 AM               147 simple.std
> 12/16/2011  08:56 AM           917,504 simple.tds
> 12/01/2011  09:40 AM             2,156 simple.tpl
> 12/16/2011  08:57 AM                 8 sims
>

3) When I try to run the -mceval step, I get the error:
> C:\Junk2\simple>simple -mceval
> Error opening file simple.psv
4) It looks like simple can not create the psv file.If I create a dummy 
psv file before hand, I get the error:

> copy simple.par simple.psv
> simple -mceval
> Incorrect value for nvar in file should be 2 but read 1968054307

5)Has anyone had experience with this?

Cheers!


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: simple.tpl
URL: <http://lists.admb-project.org/pipermail/users/attachments/20111216/5d2edb2f/attachment.ksh>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: simple.dat
Type: application/ms-tnef
Size: 171 bytes
Desc: not available
URL: <http://lists.admb-project.org/pipermail/users/attachments/20111216/5d2edb2f/attachment.bin>