Trouble with outlie() using calibrated data

[Laura Franklin]

Hi all,

I am getting this error message when trying to run the outlier function on calibrated data:

Error in x[SUB2] <- 4 * sqrt((x.SUB - 2)/(4 - x.SUB)) :
  NAs are not allowed in subscripted assignments
In addition: Warning message:
In dr > 0 & error > 0 :
  longer object length is not a multiple of shorter object length

I do not have this problem when I use data that has not been calibrated. 

I sifted through some previous conversation about this and I saw something about negative eigenvalues that could be producing the NAs - not sure if that's my case here. I've tried two functions (dplyr and base r) to convert NAs to zeros, but alas, it does not seem to make the NAs disappear (perhaps I'm embedding that function incorrectly?).

Any ideas?

Attaching my calibrated telemetry object. For context, I'm working with white-tailed deer in a dog hunt area, so it's important that I can identify outliers associated with chases. I also have an interesting phenomenon that I have not seen elsewhere in literature - multiple females making multiple excursions >1 km in every biological season (including seasons without dog hunting).

[Laura Franklin]

Whoops - I should probably include data from a moving animal :) Please ignore previous attachment, those are my actual callibrations. 

[Laura Franklin]

Alrighty, I have figured out the problem. My calibration data included the collar's fix type (3D), but I had previously dropped that column from my movement data when I cleaned it several months ago. Therefore, when I calibrated the movement data, the final result included a 3D class and an "all" class that gave estimates of 0 and CIs of infinity. All I needed to do was add a 3D column to the original movement data and the problem was solved. 

Cheers!