Missing sampling occasions in CMR models
Chloé Nater
I have time-dependent capture probabilities p[t]. Of the total 117 sampling occasions, 6 have actually seen no sampling, e.g. no individuals have been captured.
The affected occasions are: 6, 7, 8, 16, 17, 18, 38.
How do I communicate to JAGS that recapture probability p at these specific occasions = 0?
(I am aware that I could break the loops for p[t] up to manually define the affected p's as 0, but I am sure there is a more practical solution).
Thanks alot for any help!
Chloé
Schaub Michael
The easiest way may be to apply a linear model on the recapture probability. You can create a vector X in such a way that the occasion without capture effort gets the same number (e.g. 40). Thus:
X = [1, 2, 3, 4, 5, 40, 40, 40, 6, 7, ...]
The model for p then gets something like:
P[t] <- beta[X[t]]
Priors:
Beta[40] <- 0
Beta[1...39] ~ dunif(0, 1)
Best wishes
Michael
--
*** Three hierarchical modeling email lists ***
(1) unmarked: for questions specific to the R package unmarked
(2) SCR: for design and Bayesian or non-bayesian analysis of spatial capture-recapture
(3) HMecology (this list): for everything else, especially material covered in the books by Royle & Dorazio (2008), Kéry & Schaub (2012), Kéry & Royle (2015)
---
You received this message because you are subscribed to the Google Groups "hmecology: Hierarchical Modeling in Ecology" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
hmecology+...@googlegroups.com.
To post to this group, send email to
hmec...@googlegroups.com.
Visit this group at https://groups.google.com/group/hmecology.
To view this discussion on the web visit
https://groups.google.com/d/msgid/hmecology/b580f3fb-3069-4cd1-a4a2-e4ff154d5ce2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Kery Marc
wouldn't the easiest be to simply fill in dummy occasions with only NA's ?
Best regards --- Marc
Sent: 22 December 2016 16:55
To: Chloé Nater; hmecology: Hierarchical Modeling in Ecology
Subject: AW: Missing sampling occasions in CMR models
Schaub Michael
I think this is already the case. But then you have to tell the model that the p at specific occasions are 0, and that is what the sketched cosed is doing. An alternative would be actually to take these occasions out from the data set, and then no p Needs to be fixed. But then one has to deal with the unequal capture intervals when modeling survival.
Best wishes
Michael
________________________________________
Von: Kery Marc
Gesendet: Freitag, 23. Dezember 2016 12:34
An: Schaub Michael; Chloé Nater; hmecology: Hierarchical Modeling in Ecology
Betreff: RE: Missing sampling occasions in CMR models
To post to this group, send email to hmec...@googlegroups.com<mailto:hmec...@googlegroups.com>.
--
*** Three hierarchical modeling email lists ***
(1) unmarked: for questions specific to the R package unmarked
(2) SCR: for design and Bayesian or non-bayesian analysis of spatial capture-recapture
(3) HMecology (this list): for everything else, especially material covered in the books by Royle & Dorazio (2008), Kéry & Schaub (2012), Kéry & Royle (2015)
---
You received this message because you are subscribed to the Google Groups "hmecology: Hierarchical Modeling in Ecology" group.
To post to this group, send email to hmec...@googlegroups.com<mailto:hmec...@googlegroups.com>.
Kery Marc
but if all entries in the data are NA for an occasion, then I think you don't need to do anything else. p is simply not estimable and the posterior draws will be from the prior and you ignore them. I would think that you don't even need to fix the associated p's at zero or do anything else.
Best regards --- Marc
________________________________________
From: Schaub Michael
Sent: 23 December 2016 13:25
To: Kery Marc; Chloé Nater; hmecology: Hierarchical Modeling in Ecology
Schaub Michael
Yes, you are right - I haven't read this carefully enough - my answer refers to the case when there are 0 (not NA).
Best wishes
Michael
________________________________________
Von: Kery Marc
Kery Marc
thanks for confirming. So: the "least-cost" solution to unequal sampling intervals in CJS, MS and related models is then to fill in dummy occasions filled with all NA's and no other thing needs to be done, right ?
Best regards --- Marc
________________________________________
From: Schaub Michael
Sent: 23 December 2016 13:48
