Nest survival and RMARK

posts related to the RMark library, which may not be of general interest to users of 'classic' MARK

Nest survival and RMARK

Postby caspar » Tue Sep 18, 2007 9:08 am

Hi all,

Is there any particular reason why Nest survival models are not implemented in RMARK?
I have been adopting the code from functions make.mark.model and run.mark model in the package, and managed to successfully run some models while using identity matrix and logit or sin links.
However, when i try to repeatedly run the same code for different samples
i sometimes get:
"
forrtl sever (161): Program exception -array bounds exceeded
"
and a whole bunch of "Mark.exe" 's and a lot of other information that seems Greek to me.

Does any one know what happens?
I suspect that this is one of the reasons why nest survival models are not (yet?) implemented in RMARK.
I can send example code and datasubset for anyone interested in testing

Any help is much appreciated
Many thanks

Caspar
caspar
 
Posts: 14
Joined: Sun Jul 22, 2007 7:33 am

Postby caspar » Tue Sep 18, 2007 10:26 am

OK... i figured out its not a RMARK problem, nor with the code etc. Mark fails to run this particular datasets alltogether. i.e. resulting in "Program exception - array bounds exceeded"

Anyone knows what this means? Or how to overcome?

Thanks

Caspar
caspar
 
Posts: 14
Joined: Sun Jul 22, 2007 7:33 am

Postby jlbrown » Thu Sep 20, 2007 4:07 pm

If you're excited to run nest survival models in R, why not just implement a "Shaffer model" and do the analyses with GLMs? The correct link function is built in to the latest R versions, so the modeling process is about as difficult as performing any other regression. The form of the data needs to be a little different; other than that, it's a quick job. Program MARK is not needed for the GLM modeling approach. There are pros and cons to each approach; look in the recent ornithological literature for several papers addressing this issue.
jlbrown
 
Posts: 3
Joined: Wed Sep 19, 2007 5:07 pm
Location: University of Nevada, Reno

Nest survival in RMark

Postby jlaake » Sun Oct 07, 2007 1:21 pm

With Jay Rotella's help I've coded the Nest survival model into RMark and in the process of testing and writing up some help. So far appears to work fine. It has a very different structure than most MARK models so it has taken some specialized coding.

My guess about the error you encountered would be that you have some of the recorded days that are greater than the number of occasions. Don't know for sure but MARK probably sets up an array based on number of occasions and then if you had a day value outside that limit it would give the array exception. I'll include that check into the RMark code.
jlaake
 
Posts: 1417
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA


Return to RMark

Who is online

Users browsing this forum: No registered users and 15 guests