Supplement 1. A compressed package of scripts, functions, sample data, and instructions required to implement the state–space model described in the text.

dataset

posted on 2016-08-05, 10:57 authored by Greg A. Breed, Ian D. Jonsen, Ransom A. Myers, W. Don Bowen, Marty L. Leonard

File List

SSM PACKAGE Feb 2009.zip -- A package containing all needed files, including instructions, implement to models described in this paper.

Listed below are the individual files contained in SSM_PACKAGE_FEB_2009.zip:

SSM_instructions.pdf -- A set of instructions for implementing the SSM for Argos tracking data using the files provided here. Detailed descriptions of each file and how to use them may be found in this document.

.RData -- An R workspace with all scripts and functions needed already loaded.

1diff2stateM.bug -- The WinBUGS CRW state–space model file.

add.missing.dates.R -- Small subroutine for handling days with no Argos locations.

calcj.R -- Small subroutine for indexing irregular data to regular timesteps.

dat2bugslite.R -- Major subroutine for data preparation.

find.missing.dates.R -- Small subroutine for handling days with no Argos locations (needs add.missing.dates.R).

prepdat.R -- Function called to select, extract, and prepare data for WinBUGS from the sample data set testdata.csv.

runSSM.R -- Simple script that allows for easy adjustment of important MCMC parameters and executes the call to WinBUGS via wbs.R.

saveresults.R -- Function that saves the means and medians of lat, long, and behavioral state as a small text file for easy import into a mapping program of the user's choice for inspection.

seald.R -- Small subroutine that extracts raw data from the testdata.csv datafile.

step.time.R -- Small subroutine needed to index the irregular data to regular timesteps.

testdata.csv -- A sample data set including three complete grey seal tracks from the North Atlantic.

wbs.R -- The main function which calls WinBUGS from R; includes all the information to create MCMC initials.

Description

SSM_PACKAGE_Feb_2009.zip contains all scripts, functions, and sample data needed to fit the state–space correlated random walk models presented in this paper. Following the instructions (SSM_instructions.pdf) should allow readers to reproduce the results and/or fit their own Argos tracking data.