Page 1 of 1

predictDsurface, openCR, and masks

PostPosted: Wed Sep 25, 2019 8:58 am
by amurphy
Hello,

I'm trying to estimate carnivore density and see how it varies by distance to forest edge. I have six years worth of data. I'm using openCR.

Issue 1: I've attempted to create a mask list and have different masks for each year, but I've been getting an error that says that the model will just use the first mask in the list. I've been assuming that's normal, but now that I've been getting this other error, I'm thinking that this is possibly the problem.

Issue 2: So, with just one mask (due to Issue 1), I'm running a JSSASECRD analysis on the six years of data. I've created the single mask with the covariate distance to forest edge and attempted to have density vary by that. The problem is that the model doesn't seem to recognize that the mask covariate is there and keeps returning an error saying that it can't find it.

I've just troubleshooted by running a single session model in secr with the one mask and it works perfectly, finds the mask covariate, runs the model, I'm able to use predictDsurface to see how density varies with distance to forest edge.

My questions are:
1) Is the predictDsurface something that you're only able to use in secr and is that what's causing the problem?
2) Is Issue 1 (i.e., using just the first mask in a list of masks) normal or have I been overlooking a problem?

Thanks!

Re: predictDsurface, openCR, and masks

PostPosted: Wed Sep 25, 2019 4:55 pm
by murray.efford
The present version of openCR does not model density surfaces. This is stated in the vignette (section 12.5). 'secr' should do the job, and is better supported.