BayesMRA
Bayesian Multi-Resolution Gaussian Process Approximations
Software for fitting sparse Bayesian multi-resolution spatial models using Markov Chain Monte Carlo.
BayesMRA
An R package for Bayesian spatial and spatio-temporal regression using
sparse multi-resolution matrices. This package includes functions for
setting up the multi-resolution process and a function mcmc_mra() for
fitting a spatial model using Markov Chain Monte Carlo. For details
about the model and an example using simulated data, see
vignette("BayesMRA").
Installation
Package Installation
If this is your first time installing this package, make sure you have the appropriate compiler.
Installation from gitHub using devtools
To install the package from gitHub, make sure you have the appropriate
compiler. Next, make sure
you have the devtools package installed. You can install the
devtools package using
install.packages("devtools")
Once the devtools library is installed, you can install the BayesMRA
library using
devtools::install_github("jtipton25/BayesMRA")
Note: It is recommended to regularly check for updates by using
devtools::install_github("jtipton25/BayesMRA") frequently as this
package is in active development and regularly undergoes large changes
Installation from CRAN
Installation from CRAN is not currently supported (hopefully coming soon).
Make sure you have the appropriate compiler
-
Linux: It is assumed on Linux systems, the appropriate compiler is available
-
Windows: For Windows, download and install RTools. For R (>=4.0.0) follow the instructions here – older versions of R follow the instructions here.
-
MacOS: If you are on a Mac, make sure you have an openMP supported compiler – see here for instructions on how to get this setup. Follow the instructions for your specific version of R