In mhpob/TelemetryR: Handle VEMCO detection data.

Introduction and disclaimers

This code was written so that I could quickly and easily send any detected VEMCO transmitters to those who have provided their tag IDs to the ACT database. It was orignally just a script, but I've tried to transform everything into a more user-friendly function. I'm definitely not an expert coder, so these functions were written in a manner that's very specific to how I've organized things. I haven't tested any of this out on a Mac, so I'm not sure if it will work.

Hopefully, I'll be able to figure out how to make things more generic, but for now I'll use this vignette to decribe how I have things set up. If anything doesn't work, or you come up with any ideas, please feel free to let me know.

Lastly, I know naming a package "TelemetryR" is pretty presuming, but it's just what I called the scripts on my computer when I first started working on them.

Data organization

The way that I, personally, have taken to storing my detection data has been on a server owned by CBL. Since an off-site backup takes place nightly and I can access the data via the internet, this has been more useful than leaving everything in the folder created by the VUE software.

On this server, I have a folder which contains all detections, separated by arrays. Note that I have the ACT active transmitter sheet, in CSV format, in the main folder...

...and the files spit out by VUE in each array's subfolders.

Utilizing `vemsort`

What vemsort does is take the directory you provide (P:/OBrien/Biotelemetry/Detections in the pictures above) and looks for all .csv files within all subfolders, assuming that they will be the .csv files provided by VUE. It does a bit of manipulation, including converting the UTC date and time to EST/EDT and pulling out the tag standard. I also prefer working with lower-case names in R, so the columns are renamed in the process.

library(TelemetryR)
vemsort('p:/obrien/biotelemetry/detections')

I can also put in a vector of false detections that I've gotten confirmed by VEMCO. Note the difference in the length of the data frame:

vemsort('p:/obrien/biotelemetry/detections', c('A69-1601-37119','A69-1601-64288'))

If you want to go on and play with the detection data, you can pull out your own detections and stop here.

Splitting things using `ACTsplit`

ACTsplit leans on vemsort to load all of the detections into R. Because of this, there are a few more arguments to the function: those that go to vemsort and those specific to ACTsplit. You can also provide ACTsplit with an object previously created by vemsort. This is useful if you need to do more manipulation than vemsort offers before you disperse your detections to ACT. Note that the function also spits out a data frame of detections not identified by the ACT data base.

ACTsplit('p:/obrien/biotelemetry/detections')
# These are unidentified detections.

ACTsplit will create divvied up .csv files and place them in the folder you specified as the directory.

Done!

A couple of tweaks that may be needed

If you don't want the .csv files to be written in your working directory, you can choose write = F or specify a different output folder in the out argument.

ACTsplit(directory = 'p:/obrien/biotelemetry/detections',
         write = F, out = 'c:/users/secor lab/desktop')

You can also provide false and personal detections and give a date range for which detections you want (make sure the date is in year-month-day format).

# Splits detections from June 1 up to and including August 1, 2014
ACTsplit(directory = 'p:/obrien/biotelemetry/detections',
         my.trans = paste0('A69-1601-',seq(25434,25505)),
         false.det = c('A69-1601-37119','A69-1601-64288'),
         start = 20140601, end = 20140801)