[R-sig-Geo] compare UD overlap of different home range methods for the same animal

Andrew Duff andrewaduff at gmail.com
Tue Dec 21 05:23:46 CET 2010


I have what I think is a simple question for producing kerneloverlap indices
in the adehabitat package but I am having trouble with a solution.  The
basic problem is to use kernel overlap indices to compare different home
range techniques for a given animal.

 

I think the solution is in understanding how to create my own object of
khrud that contains UD's generated with different h parameters or home range
methods so that I can quantify how the different home range methods compare.
I have put together the following code from the samples using my own data,
but the solution for my problem seems to be in creating my own khrud object
containing UD's generated using different methods that I can feed into the
kerneloverlaphr function????  Maybe there is another solution?

 

Can anyone help?  I have provided a simple example to start.

 

Thanks,

 

Andy

 

#######################################################

library(adehabitat)

 

elk_tele_all <- read.table("ElkTelemetry_20100917_SPS_LE_660Feet_Final.txt",
header= TRUE, sep = ",")

elk_tele_all

names(elk_tele_all)

elk_tele_test <- subset(elk_tele_all, ElkName == 'Delores' | ElkName ==
'Gertrude' | ElkName == 'Clover' | ElkName == 'Eunice')

 

##Get the locations and animal names

loc <- elk_tele_test[, c("X_SPS", "Y_SPS")]

id <- elk_tele_test[, "ElkName"]

 

## Estimation of UD for the four animals

(ud <- kernelUD(loc, id, h = "href"))

image(ud) ## Note that the contours

 

## Example of estimation using LSCV

udbis <- kernelUD(loc, id, h = "LSCV")

image(udbis) ## Note that the contours

 

##This works to compare the four animals

kerneloverlap(loc, id,

grid=200, meth="VI", conditional=TRUE)

 

## Identical to

kud <- kernelUD(loc, id,

grid=200, same4all=TRUE)

kerneloverlaphr(kud, meth="VI", conditional=TRUE)

 

##But how do I compare the href and LSCV methods on a per animal basis?

##Once I can figure out this simple method, then I can work on more advanced
applications of CVh, BBMM, HPI, etc

???????

#######################################################

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20101220/6765a163/attachment.html>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: ElkTelemetry_20100917_SPS_LE_660Feet_Final.txt
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20101220/6765a163/attachment.txt>


More information about the R-sig-Geo mailing list