Re: [unmarked] Fixing detection at 100% with colext

[Marc Kery]

Dear Chris,

I assume that you don't have any replicate observations per site-year combination, right ?

1. Unfortunately, in unmarked there is no way to fix a parameter at a certain value such as 0 or 1. (Personally, I'd think this would be quite useful, but of course I have no clue about how hard this would be to implement !). You can, however, trick unmarked into fixing p at 1 by duplicating your detection data, so that instead of having a 0 you then have 0-0 and instead of a 1 you have 1-1. Since you then never have a "mixed" detection history per site-year combo, this makes unmarked believe that detection is perfect. You will change likelihood values etc, but for inference relative changes in the likelihood are all that counts, therefore,  I don't think that this is a problem.
2. To be honest, I am not sure how unmarked deals with missing primary periods in the context of this model fitting function. You could try to supply the data set with all-NA data filled in at the missing years and that would probably work (it would definitely work with a Bayesian analysis with JAGS and NIMBLE).

Best wishes -- Marc

---

From: unma...@googlegroups.com <unma...@googlegroups.com> on behalf of cscjsm...@gmail.com <cscjsm...@gmail.com>

Sent: Wednesday, October 8, 2025 07:52
To: unmarked <unma...@googlegroups.com>
Subject: [unmarked] Fixing detection at 100% with colext

 

Dear Unmarked community,

    First off, thanks for making such an awesome package that makes it easy for field ecologists to get meaningful answers to our questions.

       I am trying to analyze an occupancy 4 year dataset on Daurian Pika from Mongolia (a species with few published studies) with data from 2019, 2021, 2022, and 2025. We measure occupancy by checking every burrow entrance within a plot for fresh green scat, and have estimated detection probability around 100%. My 2 questions are:

1) Using function colext, is there a way to fix detection probability at 1, and does this interfere with estimation of other parameters?

2) Does the gap in years present a problem?

Thanks, 

Chris

--
*** Three hierarchical modeling email lists ***
(1) unmarked (this list): for questions specific to the R package unmarked
(2) SCR: for design and Bayesian or non-bayesian analysis of spatial capture-recapture
(3) HMecology: for everything else, especially material covered in the books by Royle & Dorazio (2008), Kéry & Schaub (2012), Kéry & Royle (2016, 2021) and Schaub & Kéry (2022)
---
You received this message because you are subscribed to the Google Groups "unmarked" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unmarked+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/unmarked/c8f855f3-6bec-4a6b-9ef5-fe9d02dd90f8n%40googlegroups.com.