Mark-resight error message

[mjordan]

I have a set of three years of camera trapping data of fishers. I am trying to run the Poisson log-normal mark-resight model on the data, and every time I run it, I get the error:

forrtl: severe (161):Program Exception - array bounds exceeded

I've used the PIMs to set up the fully time-dependent model and a model with all variables constant, and I get the same error in both cases. My input file is attached.

Other posts in this forum suggest that data sparseness might account for this error, however when I run this analysis with the example data file (Poisson_RobustDesign_OneGroup.inp), I get the same error.

I have tried uninstalling and reinstalling the latest version of MARK. Here is my hardware/software configuration.

OS: Mac OS 10.5.8 running Windows XP SP 3 in Parallels Desktop 5 (Build 5.0.9344)
MARK_INT.EXE 05/13/2010 07:10:30 PM
MARK.EXE 05/13/2010 08:46:26 PM

I also tried to do this on another Mac that has an older version of Parallels and had the same problem. I haven't tried it on a PC yet, and I'm wondering if it's a problem with the virtualization software rather than something with the input file or the way I've set up the models. Any thoughts?

Thanks!
Mark

[bmcclintock]

Hi Mark,

If the example file isn't working, I suspect perhaps the analysis hasn't been specified correctly (if starting from scratch with the .inp file) or it's a configuration issue. Are you able to run any models when opening the example results file "POISSON_ROBUSTDESIGN_ONEGROUP.DBF"? If so, then the analysis has not been specified correctly when starting from scratch. Otherwise, it sounds like a configuration issue, but I was unable to find your .inp file to verify. Could you send it to me off-forum?

Cheers,
Brett

[mjordan]

A quick follow-up to anyone else who might be following this thread ...

I got the same error message using the sample results file, and Brett was able to run the analysis successfully with my input file. This seems to be a configuration problem with my setup, and it may have to do with running XP on a Mac.

[gwhite]

I have a set of three years of camera trapping data of fishers. I am trying to run the Poisson log-normal mark-resight model on the data, and every time I run it, I get the error:

forrtl: severe (161):Program Exception - array bounds exceeded

I've used the PIMs to set up the fully time-dependent model and a model with all variables constant, and I get the same error in both cases. My input file is attached.

Other posts in this forum suggest that data sparseness might account for this error, however when I run this analysis with the example data file (Poisson_RobustDesign_OneGroup.inp), I get the same error.

I have tried uninstalling and reinstalling the latest version of MARK. Here is my hardware/software configuration.

OS: Mac OS 10.5.8 running Windows XP SP 3 in Parallels Desktop 5 (Build 5.0.9344)
MARK_INT.EXE 05/13/2010 07:10:30 PM
MARK.EXE 05/13/2010 08:46:26 PM

I found a bug that I had introduced a few weeks ago, got it fixed, and have put the new code on the MARK web site. The example file should now work.

Gary

[mjordan]

I've been in the field, so apologies for the delayed response. That bug change fixed the problem, and I was able to analyze both the example file and my own data. Thanks so much for your prompt help!

Mark