Learn R Programming

bmk (version 1.0)

MCMC diagnostics package

Description

MCMC diagnostic package that contains tools to diagnose convergence as well as to evaluate sensitivity studies, Includes summary functions which output mean, median, 95percentCI, Gelman & Rubin diagnostics and the Hellinger distance based diagnostics, Also contains functions to determine when an MCMC chain has converged via Hellinger distance, A function is also provided to compare outputs from identically dimensioned chains for determining sensitivy to prior distribution assumptions

Copy Link

Version

Install

install.packages('bmk')

Monthly Downloads

13

Version

1.0

License

GPL (>= 2)

Maintainer

Matthew Krachey

Last Published

October 17th, 2012

Functions in bmk (1.0)

Hellinger distance between distributions

Hellinger distance between two MCMC chains using a specified grid size.

Hellinger distance between two MCMC chains for sensitivity studies

MCMC.one is an mcmc object resulting from the following code:

MCMC.three is an mcmc object resulting from the following code:

Hellinger distance within consecutive batches of MCMC samples.

MCMC.one.mean0 is an mcmc object resulting from the following code:

MCMC.two is an mcmc object resulting from the following code:

MCMC.one.mean1 is an mcmc object resulting from the following code:

bmkcoverge: Convergence via the Hellinger distance

Hellinger distance between two MCMC chains using default grid in kernel density estimator.

Summaries of MCMC chains