Problems estimating overlap

[Adam Mortensen]

Hello Everyone,

I'm having projection issues while trying to estimate overlap of UDs estimated using the CTMM workflow. Many of the threads I have read along with the CTMM vignette use the buffalo data that comes with CTMM as the example. I've been having trouble adjusting those solutions to solve my problems. I am reading in location data as a csv and converting to telemetry objects as described in the CTMM workflow. Each animal is filtered from the master sheet and estimated separately. 

I'm am trying to compare the overlap of two utilization distributions that are estimated from the same animal. The first UD is estimated from a complete GPS data set from a collared fisher. The second UD is estimated from the same set of locations that I have randomly subsampled to be "VHF quality". My initial goal is to estimate the overlap of the GPS and "VHF" versions of each animal to verify that sampling method will have minimal impact on UD estimation when using CTMM. 

I've tried creating a list of UD objects from the estimated UDs to feed to the overlap function and get the error "Error in check.projections(object) : Inconsistent projections." When I check the projections I can see that while they are using the same datum, the projections are slightly different. Is there a way to assign a projection either in the original CTMM workflow or now after both UDs have been estimated? 

UD_list <- list()
UD_list$F1763_IM <- UD1763_IM_Screen_pHREML_w
UD_list$F1763_VHF <- UD1763_VHF_pHREML_w

overlap(CTMM_list, method = "Bhattacharyya", level = 0.95, debias = TRUE)
Error in check.projections(object) : Inconsistent projections.

Thank you,

Adam Mortensen

[Tyler Hodges]

Hello Adam,

When you are creating your telemetry objects via as.telemetry(), there are datum and projection arguments that you can manually set. Otherwise, ctmm automatically applies a projection. You can also use the projection() function post hoc to manipulate an object's spatial reference. Another thing to note in your code, however, is that you are trying to calculate overlap with an object called "CTMM_list", not "UD_list". Lastly, for the overlap function to work properly, the UDs have to be calculated on the same grid. This is easiest to accomplish by supplying an entire list of ctmm fits and telemetry objects into the akde() function instead of calculating them one-by-one. Hope this helps!

Best,

Tyler