Readme file
SERIES C
 Applied Statistics
Design and analysis of clustered, unmatched resource selection studies, by R. G. Clark and T. C. Strevens, pages 535–551;
Analysis of Bush Rat Data from Conjola Region (Section 3 of paper)
bushrat_allregions.txt : This tab separated text file contains the data on available and used points for all regions
# Variables on bushrat_allregions.rdata:
Site: conj, curr, jer, par (4 sites)
Point: unique identifier for point; numbers refer to used points, number+letter refers to available point selected from a regular grid
source: equal to 1 for used points, and 0 for available points
spool: an identifier for the animal for used points; negative values indicate available points
dummy: equal to 1
Logs, Leaf, Ground.Veg, Shrub.Veg, Branches: variables for the characteristics of the point: see Clark and Strevens (JRSS-C, 2008) for details
Logs1, Logs2, Leaf1, etc: indicator variables indicating level of Logs, Leaf, Ground.Veg or Shrub.Veg
eda_conj.r : This R program loads bushrat_allregions.txt and conducts some exploratory analysis including
Figure 1 from the paper.
build_model_conj.r : This R program loads bushrat_allregions.txt and fits models to it using robust and non-robust variance estimators.
The program produces Tables 2, 3 and 4 from the paper.
Simulation Study: Section 4 of paper
generate_grid.r : This R program generates the 1000 by 1000 grids of available points, and also generates random paths through the grid for
several populations of animals.
simulate.r : This R program simulates sampling of used and available points from the populations generated by generate_grid.r
summarise.r : This R program summarises the results from simulate.r, including Tables 5-8 and Figure 3 from the paper.
##########
Notes
##########
If the programs are downloaded and run, the user will need to change the directory names in the programs to the appropriate directory on their computer.
A number of R libraries are used, these are indicated at the start of each program. All libraries are available from www.r-project.org .
You are welcome to use these programs at your own risk. Some of the programs use a large amount of memory, and may cause instability or crashes.
The authors accept no liability for errors or problems. Keep your machine secure and always use anti-virus and anti-spyware software!
Robert G. Clark
Centre for Statistical and Survey Methodology
School of Mathematics and Applied Statistics
University of Wollongong
Wollongong
NSW 2522
Australia
E-mail: rclark@uow.edu.au
|
