nonparboot for confidence intervals of derived parameter

[Tyler Hallman]

Hi Everyone,

I'm having a bit of trouble figuring out how to use the nonparboot function in the unmarked package to calculate confidence intervals around a derived parameter. Specifically, I am using a distsamp model to estimate density throughout a landscape. I want to use these densities to calculate a derived population estimate for the entire landscape. I'm comparing the estimates to those from other modeling methods in which I use non-parametric bootstrapped confidence intervals. 

Unlike the parboot function, which has a very clear help file, the nonparboot doesn't. From examples, it seems that the arguments included for nonparboot are only the model object and the number of iterations. For a derived parameter such as landscape level population estimates, I would need to include a function for the statistic of interest. Can this be done in the same way it's done for the parboot function? From the text in the help file it appears that I would only have the bootstrap samples added on and I would then need to loop through this object to get what I'm going for. 

Any help would be greatly appreciated. I have some experience with the boot function in the boot package, so I could always use that. 

Cheers,

-Tyler Hallman

[Richard Chandler]

Hi Tyler,

The nonparboot function is pretty limited and can't do what you're after. Perhaps 'boot' is the way to go. 

Richard

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Richard Chandler

Assistant Professor

Warnell School of Forestry and Natural Resources

University of Georgia

706-542-5815

chandlerlab.uga.edu