Performs reversible-jump Markov chain Monte Carlo (Green, 1995) <doi:10.2307/2337340>, specifically the restriction introduced by Barker & Link (2013) <doi:10.1080/00031305.2013.791644>. By utilising a 'universal parameter' space, RJMCMC is treated as a Gibbs sampling problem. Previously-calculated posterior distributions are used to quickly estimate posterior model probabilities. Jacobian matrices are found using automatic differentiation.
| Version: | 0.4.4 |
| Depends: | madness, R (≥ 3.2.0) |
| Imports: | utils, coda, mvtnorm |
| Suggests: | knitr, FSAdata |
| Published: | 2019-03-02 |
| Author: | Nick Gelling [aut, cre], Matthew R. Schofield [aut], Richard J. Barker [aut] |
| Maintainer: | Nick Gelling <nickcjgelling at gmail.com> |
| License: | GPL-3 |
| NeedsCompilation: | no |
| Materials: | README |
| CRAN checks: | rjmcmc results |
| Reference manual: | rjmcmc.pdf |
| Vignettes: |
R Package rjmcmc: The Calculation of Posterior Model Probabilities from MCMC Output |
| Package source: | rjmcmc_0.4.4.tar.gz |
| Windows binaries: | r-devel: rjmcmc_0.4.4.zip, r-release: rjmcmc_0.4.4.zip, r-oldrel: rjmcmc_0.4.4.zip |
| OS X binaries: | r-release: rjmcmc_0.4.4.tgz, r-oldrel: rjmcmc_0.4.4.tgz |
| Old sources: | rjmcmc archive |
Please use the canonical form https://CRAN.R-project.org/package=rjmcmc to link to this page.