Bug in make.capthist for multi detectors

Richard Glennie

unread,

Jan 29, 2021, 7:15:48 PM1/29/21

to secr

Hello all,

Just wanted to query whether this is a possible bug in secr 4.3.3 or I am misunderstanding something.

I found some unusual behaviour when using make.capthist with multi detectors. It reported a warning that it was "dropping repeat detections within occasions at exclusive detectors (traps)" when I was sure I had removed all such repeat detections.

Find attached a script that reproduces the problem for a minimal example and proposes where I think the problem is in make.capthist.

Best,

Richard

scr_multi_repeated.R

Murray Efford

unread,

Jan 30, 2021, 11:33:07 AM1/30/21

to secr

Hi Richard

It's a bug, of course. Thanks for your detailed report. I have fixed it in the pre-release version available here secr_4.3.4.zip (Windows binary) or here secr_4.3.4.tar.gz (source).

As a workaround, you can also read data as if from a "proximity" detector and then convert to "multi":

traps<- make.grid(detector = "proximity")
chp <- make.capthist(captures, traps)
chm <- reduce(chp, outputdetector = "multi", dropunused = FALSE)
The result is not identical but the differences are subtle and I think ignorable.

Murray

Richard Glennie

unread,

Feb 1, 2021, 5:54:57 AM2/1/21

to secr

Hi Murray:

Thanks for the rapid reply and quick code turnaround. I ended up appending a letter to the end of the ID's and that seemed to work.