model.average in POPAN - strange phi values

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

model.average in POPAN - strange phi values

Postby SoConfused » Tue Nov 07, 2017 7:45 am

I'm running a POPAN model on a dataset with 3 variables - Var1 (11 levels, nested in Var2), Var2 (2 levels), and Var3 (2 levels, crossed). When I estimate phi using model.average, I get a strange "extra" phi value. This behaviour occurs even if there's a single model being run, but I can't reproduce it using the dipper data, even if I add simulated variables.

Here's the model setup:
Code: Select all
data1.proc <- process.data(data1, model="POPAN", begin.time = 2000, groups = c("Var3", "Var2", "Var1"))
data1.ddl <- make.design.data(data1.proc)

data1.analysis=function(){   
   Phi.Var1 = list(formula= ~ -1 + Var1,link="sin")
   pent.1 = list(formula = ~1)
   p.Var3 = list(formula= ~ Var3)
   N.Var2.x.Var3 = list(formula = ~ Var2*Var3)
   cml=create.model.list("POPAN")
   mark.wrapper(cml,data=data1.proc, ddl=data1.ddl)}
 
data1.results <- data1.analysis()


To look at results, I did this, which gave me a total of 12 values. The majority of phis was unique to Var1. However, a single level of Var1 had two different values of phi produced.
Code: Select all
temp <- model.average(data1.results, "Phi", vcv=TRUE)$estimates
temp <- unique(select(temp, estimate, lcl, ucl, Var1))
temp <- unique(temp)
dim(temp)


This did not happen if I simply looked at
Code: Select all
data1.results[[1]]$results$real
- this only produced a single value of phi for each level of Var1. I could post the data here (if I figured out how) if that helps.

I wanted to use model.average, since my original code includes many models. What could be the problem here?

Thank you!
SoConfused
 
Posts: 53
Joined: Wed Nov 05, 2014 8:25 am

Re: model.average in POPAN - strange phi values

Postby jlaake » Tue Nov 07, 2017 9:46 am

Please try this without sin link to see if the same thing occurs.
jlaake
 
Posts: 1417
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: model.average in POPAN - strange phi values

Postby SoConfused » Tue Nov 07, 2017 10:56 am

It doesn't converge without the sin link - I'm guessing it's not happy about survivals being too close to 1.
SoConfused
 
Posts: 53
Joined: Wed Nov 05, 2014 8:25 am


Return to RMark

Who is online

Users browsing this forum: No registered users and 19 guests

cron