An R package to conveniently run JAGS in parallel from R
jagstools consists of one main function, jags_sample that is a wrapper for
functions in the rjags package that makes it easier to sample MCMC chains
using parallel processing. Parallel processing is conducted by the parLapply
function in the parallel package. Unlike other packages that have similar
functionality (e.g., R2jags and runjags) jags_sample returns an
mcmc.list object that is easily manipulated using the coda package, rather
than defining a new class. Because of this, in its current state, models fit
using jags_sample cannot be updated later by running the sampler longer.
The jagstools package is not on CRAN, but can be installed from R using:
install.packages("devtools")
library(devtools)
install_github(repo = "michaelmalick/r-jagstools")
library(jagstools)
MIT