Biodiversity is in crisis. The overarching aim of conservation
is to preserve biodiversity patterns and processes. To this end, protected
areas are established to buffer species and preserve biodiversity processes.
But resources are limited and so protected areas must be cost-effective.
This package contains tools to generate plans for protected areas
(prioritizations), using spatially explicit targets for biodiversity
patterns and processes. To obtain solutions in a feasible amount of time,
this package uses the commercial 'Gurobi' software package (obtained from
< http://www.gurobi.com/>). For more information on using
this package, see Hanson et al. (2018)
Biodiversity is in crisis. The overarching aim of conservation is to preserve biodiversity patterns and processes. To this end, protected areas are established to buffer species and preserve biodiversity processes. But resources are limited and so protected areas must be cost-effective. This package contains tools to generate plans for protected areas (prioritizations), using spatially explicit targets for biodiversity patterns and processes. To obtain solutions in a feasible amount of time, this package uses the commercial 'Gurobi' software package (obtained from http://www.gurobi.com/). For more information on using this package, see our paper published in Methods in Ecology and Evolution (http://dx.doi.org/doi:10.1111/2041-210X.12862).
This package depends on several packages which can be difficult to install under Linux and Mac operating systems.
Linux (Ubuntu) users can install them typing the following code into the terminal:
sudo apt-get update
sudo apt-get install libgdal-dev
sudo apt-get install libproj-dev
sudo apt-get build-dep r-cran-rgl
Mac OSX users can install them using the code below.
brew install Caskroom/cask/xquartz
brew install gdal
Rscript -e "setRepositories(ind = 1:2); install.packages(c('rgdal', 'rgeos'))"
To install the latest official version on CRAN, use the following R code.
install.packages("raptr")
To install the development version on GitHub, use the following R code.
if (!require("devtools"))install.packages("devtools")devtools:::install_github("jeffreyhanson/raptr")
Once this package has been installed, you can read through the vignette for a tutorial on how to use it.
View it here, or by running the R code below.
vignette("raptr", package = "raptr")
citation("raptr")
##
## To cite the raptr package in publications, use:
##
## Hanson JO, Rhodes JR, Possingham HP & Fuller RA raptr:
## Representative and Adequate Prioritization Toolkit in R. Methods
## in Ecology & Evolution, 9: 320--330. DOI:
## 10.1111/2041-210X.12862
##
## A BibTeX entry for LaTeX users is
##
## @Article{,
## author = {Jeffrey O Hanson and Jonathan R Rhodes and Hugh P Possingham and Richard A Fuller},
## title = {raptr: Representative and adequate prioritization toolkit in R},
## journal = {Methods in Ecology and Evolution},
## year = {2018},
## volume = {9},
## pages = {320--330},
## url = {https://dx.doi.org/10.1111/2041-210X.12862},
## }
##
## You may also want to cite the package version. Find it with
## "help(package=raptr)".
GurobiOpts
.print.RapResults
function now prints information in a prettier manner.raptr::solve
function now throws a warning if some species are poorly
represented in the solution.knitr::rmarkdown_notangle
engine to avoid running the
code during package checks, and placeholder vignette file has been removed.pkg::function
.message
following best practices.is.GurobiInstalled
now does not run gurobi to avoid spurious notes during
CRAN checks on academic license.make.DemandPoints
.requireNamespace
.