Derived Psi for multi-season model

questions concerning analysis/theory using program PRESENCE

Derived Psi for multi-season model

Postby nathanvs » Sun Nov 22, 2015 6:55 pm

I'm trying to get probability of occupancy (psi) values for unsurveyed sites in a metapopulation in multiple years, for use as a connectivity covariate on colonization. Because of the difficulties the alternative multi-season has converging, I'm trying to use the standard parameterization (initial psi, gamma, epsilon) and then use the derived psi estimates PRESENCE outputs to calculate my connectivity covariate... Essentially an autologistic approach, though not using PRESENCE's built-in psi1 covariate approach; I would like to calculate the covariate outside of PRESENCE. Two questions:

1- The newest versions of PRESENCE no longer seem to output derived psi values for the standard multi-season occupancy model, only derived lambda. Is this a bug, or were these removed for some reason?

2- Is there any issue with including a large number of unsurveyed sites? My additional sites are essentially blank rows ( - - - - - - ) for survey data with site-level covariate data. From my preliminary runs on older PRESENCE versions, it seems like running the model with and without the unsurveyed sites, the models outputs & AIC are identical, it just provides estimated psi values for the unsurveyed sites as well (which is what I want). This is what I would expect given that they aren't contributing any information to the likelihood function. However, I want to make sure I'm not missing anything important here.

Thank you!
Nathan
nathanvs
 
Posts: 8
Joined: Fri Nov 20, 2015 5:07 pm

Re: Derived Psi for multi-season model

Postby jhines » Sun Nov 22, 2015 10:39 pm

1. It was a bug and has just been fixed (today 11/22/2015).

2. No problem with unsurveyed sites. AIC should be the same as if the unsurveyed sites were not included.
jhines
 
Posts: 599
Joined: Fri May 16, 2003 9:24 am
Location: Laurel, MD, USA

Re: Derived Psi for multi-season model

Postby nathanvs » Tue Nov 24, 2015 5:26 pm

Hmm... I'm still not seeing derived PSI outputs on the newest verison (10.0); has this fix been uploaded yet?
nathanvs
 
Posts: 8
Joined: Fri Nov 20, 2015 5:07 pm


Return to analysis help

Who is online

Users browsing this forum: No registered users and 12 guests