Page 1 of 1
Random Effects for Modeling sites in R-Presence

Posted:
Mon Feb 08, 2021 7:06 pm
by Suarez_B
Aloha,
Forgive me if this has already been discussed but I am having difficulties trying to find random effects in RPresence, I see that Program Mark is capable but I am not finding such a thing in the RPresence package. Any help would be greatly appreciated.
Thanks in advance.
Re: Random Effects for Modeling sites in R-Presence

Posted:
Mon Feb 08, 2021 8:00 pm
by cooch
I believe that this is still the case. Or, you could go all in and code everything up in JAGS or Stan, and do it full-on Bayesian.
Re: Random Effects for Modeling sites in R-Presence

Posted:
Mon Feb 08, 2021 8:33 pm
by Suarez_B
Thank you for the speedy reply! Unfortunately, I am not up to speed in JAGS and doing my own coding at this point. I am a foundling to occupancy and population studies but quick to learn, I will definitely take a look at both of those! Thank you again
Re: Random Effects for Modeling sites in R-Presence

Posted:
Tue Feb 09, 2021 9:28 am
by cooch
So, given that, is there a particular reason you're not using MARK, which (as noted) does what you want?
Re: Random Effects for Modeling sites in R-Presence

Posted:
Tue Feb 09, 2021 6:10 pm
by Suarez_B
I used R-Presence for a report I did late last year looking at a number of tropical reef fishes so I am familiar with that program and how it works in R. I am actually looking at Package Unmarked as a solution as it does random effects as well, but I am interested in looking at the same data set using a zero-inflated negative binomial model using actual count data seeing that the data set I am working with violates a whole lot of assumptions. As I said I am relatively new to this in general.
Re: Random Effects for Modeling sites in R-Presence

Posted:
Wed Feb 10, 2021 9:50 am
by cooch
Suarez_B wrote:I used R-Presence for a report I did late last year looking at a number of tropical reef fishes so I am familiar with that program and how it works in R. I am actually looking at Package Unmarked as a solution as it does random effects as well, but I am interested in looking at the same data set using a zero-inflated negative binomial model using actual count data seeing that the data set I am working with violates a whole lot of assumptions. As I said I am relatively new to this in general.
Fair enough -- I was simply curious. unmarked has pretty strong capabilities, although there is a fair learning curve (IMO). But, their Google group offers excellent support.
Best of luck!