running same model set w/ different datasets

questions concerning analysis/theory using program MARK

running same model set w/ different datasets

Postby jlehmkuhl » Mon Jul 21, 2008 6:13 pm

Is there and easy way to run the same set of models with different datasets in MARK? I have RTFM etc and can't find anything helpful. I have field data on occurrence of several species at the same sampling points. eg. bird point counts. I want to estimate occupancy for each species with the same set of model hypotheses, but the "best" model conceivably could differ by species.

I rather not rebuild the model PIMs for each species, so I could...

a) have all the species data in one file and model with a species covariate or group var. But the extra covariate/group complicates the analysis.

or

b) is there a way to save the model set and run it with separate species' data files?

Any advice on the best way?
jlehmkuhl
 
Posts: 18
Joined: Thu Aug 21, 2003 8:04 pm

RMark is your friend -- RE: same model set, new data set

Postby dhewitt » Mon Jul 21, 2008 6:28 pm

Jeff will say it if I don't, so ...

Appendix C, section 13 -- starting on pg. 54 of the Gentle Introduction.

This is where RMark shines.
dhewitt
 
Posts: 150
Joined: Tue Nov 06, 2007 12:35 pm
Location: Fairhope, AL 36532

Postby jlaake » Mon Jul 21, 2008 6:36 pm

If you don't want to go the RMark route, as long as the model structure for the data is the same (same number of occasions, groups etc), in MARK you can start a new project and then there is a menu option: Run/File model(s) which you select from the other dbf/fpt. It copies over the design matrix and PIMS. However, if the structure changes, you are screwed. For example, if you collect another year of data, then you start from scratch. I work with a lot of monitoring programs where we are always adding a new occasion each year. Starting over with MARK is what drove me to create RMark. If you write the RMark script correctly, the entire analysis can be re-run without changing anything other than the input data.

--jeff
jlaake
 
Posts: 1480
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

same model set, new data

Postby jlehmkuhl » Wed Jul 23, 2008 2:10 pm

Thanks, all, for the suggestions. Another reason to make the leap to R.
jlehmkuhl
 
Posts: 18
Joined: Thu Aug 21, 2003 8:04 pm

Re: same model set, new data

Postby bacollier » Wed Jul 23, 2008 10:13 pm

jlehmkuhl wrote:Thanks, all, for the suggestions. Another reason to make the leap to R.


For what it is worth, I just finished a project doing exactly what you are trying to do. I found it easiest to import all the data in, then use setwd() to put the RMARK modeling output to different folders for each dataset. Only reason I am adding to Jeff's comments is that if you have a considerable model set (e.g., say 15 models for 5 datasets) weeding through the RMARK output for 15 model outputs*5 datasets*3 different files/run takes time, so I found that separating them helped.
bacollier
 
Posts: 231
Joined: Fri Nov 26, 2004 10:33 am
Location: Louisiana State University


Return to analysis help

Who is online

Users browsing this forum: No registered users and 1 guest