Utilities for Processing Posterior Samples Stored in 'mcmc.lists'

The aim of 'postpack' is to provide the infrastructure for a standardized workflow for 'mcmc.list' objects. These objects can be used to store output from models fitted with Bayesian inference using 'JAGS', 'WinBUGS', 'OpenBUGS', 'NIMBLE', 'Stan', or even custom MCMC algorithms. Although the 'coda' R package provides some methods for these objects, it is somewhat limited in easily performing post-processing tasks for specific nodes. Models are ever increasing in their complexity and the number of tracked nodes, and oftentimes a user may wish to summarize/diagnose sampling behavior for only a small subset of nodes at a time for a particular question or figure. Thus, many 'postpack' functions support performing tasks on a subset of nodes, where the subset is specified with regular expressions. The functions in 'postpack' streamline the extraction, summarization, and diagnostics of specific monitored nodes after model fitting. Further, because there is rarely only ever one model under consideration, 'postpack' scales efficiently to perform the same tasks on output from multiple models simultaneously, facilitating rapid assessment of model sensitivity to changes in assumptions.


News

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("postpack")

0.5.2 by Ben Staton, a month ago


https://bstaton1.github.io/postpack/


Report a bug at https://github.com/bstaton1/postpack/issues


Browse source code at https://github.com/cran/postpack


Authors: Ben Staton [aut, cre]


Documentation:   PDF Manual  


MIT + file LICENSE license


Imports stringr, matrixcalc, coda, mcmcse, abind

Suggests knitr, rmarkdown


See at CRAN