‘postpack’ exists to facilitate processing the output of models fitted using Markov Chain Monte Carlo (MCMC) methods, as is commonly done in Bayesian inference. Although substantial capabilities exist to interface from R with programs that perform MCMC like JAGS, WinBUGS, OpenBUGS, NIMBLE, and Stan, the functionality to easily process the output is sometimes lacking.
In particular, it is often cumbersome to perform post-processing tasks for specific nodes from the model.
‘postpack’ makes extracting specific nodes more transparent by accepting regular expressions and returning output in predictable and manipulable formats.
Common tasks encompassed by “post-processing” include:
‘postpack’ seeks to enforce consistent rules in how many these
actions are performed by the user, all based around
mcmc.list
objects (one per model).
Want to learn more?
mcmc.list
objects used
in the examples and vignettes, take a look at that
vignette.Ready to give ‘postpack’ a try in your workflow?
A stable version can be found on CRAN:
install.packages("postpack")
Or the development version can be installed via:
remotes::install_github("bstaton1/postpack")
Many thanks to Henry Hershey for designing the hex sticker and for feedback on overall usability.