Issues with N occasions and detector type in secr 4.1.0

questions concerning anlysis/theory using program DENSITY and R package secr. Focus on spatially-explicit analysis.

Issues with N occasions and detector type in secr 4.1.0

Postby charlottesearle » Sun Jan 05, 2020 12:42 pm

Hi everyone, and happy New Year!

My name is Charlotte Searle, and I’m a PhD student at the University of Oxford. I’m currently using secr to estimate population density for leopards at sites in the Ruaha-Rungwa landscape in southern Tanzania, using camera trap data I collected in 2018 and 2019.

I originally ran some of these analyses back in February 2019 in secr 3.2.0. This week I have been re-running the same analyses in secr 4.1.0, as I wanted to check the buffer widths and finalise my results, but noticed that the density estimates were coming out slightly different from those produced last year using the exact same input data.

My first thought was that there is a problem with the input files – so I checked and re-checked these – but on closer inspection, it seems that there is a problem arising when fitting the models. When I fit a model and print the output, where previously it would say "N occasions : [correct number of sampling occasions]", it now always says "N occasions : 1". I’ve tried this for four different data sets, with different numbers of sampling occasions, and the same problem is affecting all of them. However, when I print a summary of the capthist it does show the correct number of occasions, so the input files seem to be behaving as expected.

The other change I’ve noticed is that the detector type in the model output from secr 4.1.0 is listed as count, rather than proximity, despite including "detector = “proximity”" in the read.capthist function, as per my previous analyses. As above, when I print the capthist summary, it does list the detector type as proximity, so the problem seems to be arising later on.

Output of model fitted in February 2019 (secr 3.2.0):

secr.fit(capthist = RNP.Lonly.capthist, model = list(g0 ~ 1,
sigma ~ 1, D ~ 1), buffer = buffer, detectfn = "HN", binomN = 1,
hcov = "V5", details = list(minprob = 1e-200), trace = FALSE)
secr 3.2.0, 16:41:01 06 Feb 2019

Detector type proximity
Detector number 45
Average spacing 1957.079 m
x-range 680198 706553 m
y-range 9146385 9159243 m

Usage range by occasion
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
min 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
max 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
min 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
max 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83
min 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0
max 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

N animals : 37
N detections : 201
N occasions : 83
Mask area : 144114.7 ha
...


Output of same model fitted to the same data in January 2020 (secr 4.1.0):

secr.fit(capthist = RNP.Lonly.capthist, model = list(g0 ~ 1,
sigma ~ 1, D ~ 1), buffer = buffer, detectfn = "HN",
binomN = 1, hcov = "V5", details = list(minprob = 1e-200),
trace = FALSE)
secr 4.1.0, 15:11:10 05 Jan 2020

Detector type count
Detector number 45
Average spacing 1957.079 m
x-range 680198 706553 m
y-range 9146385 9159243 m

Usage range by occasion
1
min 63
max 83

N animals : 37
N detections : 201
N occasions : 1
Count model : Binomial, size from usage
Mask area : 144114.7 ha
...


If anyone has any advice on how to fix this problem I would really appreciate it!

Best,
Charlotte
charlottesearle
 
Posts: 3
Joined: Sun Jan 05, 2020 11:22 am

Re: Issues with N occasions and detector type in secr 4.1.0

Postby murray.efford » Sun Jan 05, 2020 4:48 pm

You should also notice that version 4 is much faster. See https://www.otago.ac.nz/density/pdfs/secr-version4.pdf. You can return to the old behaviour by setting details = list(fastproximity = FALSE).
murray.efford
 
Posts: 686
Joined: Mon Sep 29, 2008 7:11 pm
Location: Dunedin, New Zealand

Re: Issues with N occasions and detector type in secr 4.1.0

Postby charlottesearle » Sun Jan 05, 2020 6:09 pm

Hi Murray,

I had noticed that it is remarkably fast!

Thanks for passing on that documentation; I'm sorry I hadn't found it myself. It's good to hear that's all as intended. There's clearly something else funny going on with my input data, and I wrongly attributed the differences caused by that to the different occasions/detector type.

Sorry, and thanks for your help!

Best,
Charlotte
charlottesearle
 
Posts: 3
Joined: Sun Jan 05, 2020 11:22 am

Re: Issues with N occasions and detector type in secr 4.1.0

Postby murray.efford » Sun Jan 05, 2020 6:11 pm

No problem. I've adjusted the web page and tweaked secr-version4.pdf to make this more obvious - it's tripped up quite a few people.
murray.efford
 
Posts: 686
Joined: Mon Sep 29, 2008 7:11 pm
Location: Dunedin, New Zealand


Return to analysis help

Who is online

Users browsing this forum: No registered users and 8 guests

cron