limited time varying capture recapture probabilities

posts related to the RMark library, which may not be of general interest to users of 'classic' MARK

limited time varying capture recapture probabilities

Postby equallen » Thu Jul 21, 2022 7:45 pm

Hello,

I am new to transitioning from MARK to RMARK and am having difficulty with the model notation. I would like to run a robust design model such that capture and recapture probabilities only vary on day 1, but are constant for every day of trapping within a session after day 1. In MARK, this would like like p=c(1,2) but I am having trouble figuring out how to create a valid formula for this in RMARK.
equallen
 
Posts: 1
Joined: Thu Jul 21, 2022 6:58 pm

Re: limited time varying capture recapture probabilities

Postby jlaake » Wed Aug 10, 2022 4:38 pm

I just noticed that this post never got a reply. Sorry about that. Essentially you need to add design data that creates a factor variable which pools the days after day1 or just by having a numeric variable that is 1 for day 1 an 0 for all other days. Below is an example but if that is not what you intended you'll have to provide more information. --jeff

Here is an examples using the robust data in RMark

Code: Select all
data(robust)
# data from Robust.dbf with MARK
# 5 primary sessions with secondary sessions of length 2,2,4,5,2
#
time.intervals=c(0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0)
rb.proc=process.data(robust,model="Robust",time.intervals=time.intervals)
rb.ddl=make.design.data(rb.proc)

# create variable in p design data which is 1 for day (time) 1 and 0 otherwise
rb.ddl$p$day1=ifelse(rb.ddl$p$time==1,1,0)
# need to add the same variable to c because we are sharing p/c - in this case day1=0 because c is only for occasions past day1
rb.ddl$c$day1=0

S.time=list(formula=~time)
# p varies by day1 for each session and by session for other days but constant across those days within session
p.time.session=list(formula=~-1+session:day1+session,share=TRUE)

mark(rb.proc,rb.ddl,model.parameters=list(S=S.time,p=p.time.session))


Here is beginning of output.
Output summary for Robust model
Name : S(~time)Gamma''(~1)Gamma'(~1)p(~-1 + session:day1 + session)c()f0(~session)

Npar : 21
-2lnL: -18057.08
AICc : -18014.92

Beta
estimate se lcl ucl
S:(Intercept) 2.0771695 0.1325296 1.8174115 2.3369276
S:time2 -0.1977186 0.2091827 -0.6077167 0.2122794
S:time3 -0.9472555 0.1654696 -1.2715759 -0.6229351
S:time4 -0.8406458 0.3081152 -1.4445515 -0.2367400
GammaDoublePrime:(Intercept) -1.9649189 0.0995623 -2.1600611 -1.7697767
GammaPrime:(Intercept) -2.1762911 0.5102500 -3.1763811 -1.1762011
p:session1 0.5996324 0.0795988 0.4436187 0.7556461
p:session2 0.3431816 0.0803888 0.1856196 0.5007436
p:session3 0.5910726 0.0491665 0.4947064 0.6874389
p:session4 0.4682220 0.0471904 0.3757288 0.5607153
p:session5 -0.1260259 0.1436539 -0.4075875 0.1555357
p:session1:day1 0.1835033 0.0948776 -0.0024569 0.3694634
p:session2:day1 0.2275862 0.1031532 0.0254059 0.4297666
p:session3:day1 -0.6412770 0.0893858 -0.8164733 -0.4660808
p:session4:day1 -0.1575436 0.1007473 -0.3550082 0.0399210
p:session5:day1 0.0803370 0.1417940 -0.1975793 0.3582534
f0:(Intercept) 4.7123162 0.1427475 4.4325311 4.9921012
f0:session2 0.0771959 0.1972439 -0.3094020 0.4637939
f0:session3 -1.9772190 0.3179338 -2.6003694 -1.3540687
f0:session4 -3.2746265 0.5575347 -4.3673946 -2.1818584
f0:session5 -0.0291131 0.2471420 -0.5135114 0.4552852

jlaake
 
Posts: 1320
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA


Return to RMark

Who is online

Users browsing this forum: Google [Bot] and 2 guests