RMark covariate.predictions with mlogit

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

RMark covariate.predictions with mlogit

Postby jlaake » Thu Apr 09, 2015 10:57 am

Sometimes it doesn't pay to always get the newest version. I made a change to covariate.predictions in versions >2.1.9 that has messed up predictions with parameters like pent, Psi etc with the mlogit link. I had patched it to handle multiple individual covariate values with mlogit parameters but in doing so I messed up the case where there were no covariates. Do not use covariate.predictions for mlogit parameters unless you have one of the earlier versions. I'll get this patched soon and on CRAN and will post a message when it is fixed. Sorry for any inconvenience.

--jeff
jlaake
 
Posts: 1479
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: RMark covariate.predictions with mlogit

Postby jlaake » Thu Apr 09, 2015 11:09 am

One further point here. The function popan.derived uses covariate.predictions and it will not produce correct results in versions >2.1.9; however in 2.1.12 all of the derived parameters are now extracted from the MARK binary file so popan.derived is now much less useful.
jlaake
 
Posts: 1479
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: RMark covariate.predictions with mlogit

Postby jlaake » Thu Apr 09, 2015 6:59 pm

Version 2.1.13 of RMark is available. You can get the Windows binary or tar.gz file for building on Linux or Mac at
https://drive.google.com/open?id=0B77g1ScdUwVeVU9rUE5lVS1kZ2M&authuser=0

I'd appreciate hearing about any tests of the new version of RMark and in particular popan.derived or covariate.predictions. I've tested popan.derived and a few examples of covariate.predictions and all is being computed correctly. But any further testing would be appreciated. The changes should not affect values that don't use mlogit links. Once I'm confident that I've got it right, I'll post to CRAN.

The following is from the package NEWS on the changes made to covariate.predictions

A bug in covariate.predictions was introduced in the change for 2.1.10. Incorrect values were being computed for mlogit parameters when covariate values were not used. popan.derived used that approach and incorrect values resulted for the abundance estimates. Thanks to Leigh Ann Starcevich for reporting this discrepancy. A few other changes were made in covariate.predictions to prevent possible errors with mlogit parameters. If you are making the computations for parameters that use an mlogit link you must use the separate indices argument if you have individual covariate values. If you try to use the data=data.frame(index=...,cov=) approach with mlogit parameters and you have covariate values, the function will stop with an error. Also, if you only include a portion of the indices in an mlogit set, it will also stop and issue an error and will tell you the set of indices that should be included for that mlogit set. If you were allowed to exclude some indices the result would be incorrect because the real parameter values are computed by summing over the entire mlogit set.
jlaake
 
Posts: 1479
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: RMark covariate.predictions with mlogit

Postby murray.efford » Thu Apr 09, 2015 9:42 pm

Jeff
Congratulations for keeping RMark rolling forwards! I'm also posting out of sympathy for your mlogit issues - I've found it a beast to handle myself. My question is: what would you recommend as a single, authoritative and reasonably complete description of the mlogit tranformation as it relates to capture-recapture?
Murray
murray.efford
 
Posts: 712
Joined: Mon Sep 29, 2008 7:11 pm
Location: Dunedin, New Zealand


Return to RMark

Who is online

Users browsing this forum: No registered users and 1 guest