Freedom251 › Statistical Analysis Software

rmcmc

Free plan

What rmcmc is

rmcmc is a free MIT-licensed R package for simulating Markov chains and computing MCMC estimates of expectations for target distributions on real-valued vector spaces. Users can define a target with a log-density expression, separate density and gradient functions, or a BridgeStan StanModel when BridgeStan’s R interface is installed. Metropolis-Hastings transitions support random-walk, Langevin, Hamiltonian, and Barker proposals, with Barker as the default. Warm-up adapters can tune proposal scale and shape before the main chain; those proposal settings remain fixed during main iterations. Barker also accepts a custom auxiliary-noise sampler, and the package includes a bimodal-noise convenience function. Sampling returns the final chain state, main-iteration traces, and transition statistics. Enabling trace_warm_up adds warm-up traces and statistics. Generated chains can be passed to posterior summary and convergence-diagnostic functions, and the site describes robust adaptive Metropolis proposal adaptation. Installation is available from CRAN with install.packages("rmcmc") or from GitHub for the development version. Documentation includes a Stan workflow example using neuroscience data.

Who it suits

rmcmc suits R users who need to define target distributions and sample Markov chains with several Metropolis-Hastings proposals. It also supports users integrating Stan models through BridgeStan.

What it does well

Free MIT-licensed R package
Four Metropolis-Hastings proposal types
Warm-up scale and shape adaptation
Optional warm-up traces and statistics
BridgeStan and Stan model integration

What to know first

Requires R and package installation
BridgeStan features require its R interface
Main-chain proposal settings stay fixed after warm-up
GitHub issues are the listed bug-report route

Verdict

rmcmc offers configurable MCMC sampling with adaptation, diagnostics-ready output, and optional Stan integration. Its usefulness depends on working in R and, for Stan workflows, installing BridgeStan’s interface.

Facts from rmcmc's own pages

Adaptation: During warm-up, adapters can tune proposal scale and shape; proposal parameters remain fixed during the main chain iterations. · github-pages.ucl.ac.uk, 12 Oct 2026
BridgeStan integration: If BridgeStan’s R interface is installed, it can be used to specify the target distribution for sampling. · github-pages.ucl.ac.uk, 12 Oct 2026
Custom noise: The Barker proposal accepts a custom auxiliary-noise sampling function, and the package provides a convenience function for bimodal noise. · github-pages.ucl.ac.uk, 12 Oct 2026
Default proposal: Chains use the Barker proposal by default, a gradient-based method inspired by the Barker accept-reject rule. · github-pages.ucl.ac.uk, 12 Oct 2026
Free plan: Yes · github-pages.ucl.ac.uk, 23 Sept 2026
Installation: The site gives CRAN installation instructions and instructions for installing the development version from GitHub. · github-pages.ucl.ac.uk, 12 Oct 2026
Intended use: The package is for users who want to define a target distribution and sample a Markov chain, with a documented example fitting a model to neuroscience data using Stan. · github-pages.ucl.ac.uk, 5 Oct 2026
License: The site identifies the package license as MIT. · github-pages.ucl.ac.uk, 12 Oct 2026
Maintainer: Matthew M. Graham is listed as author and maintainer. · github-pages.ucl.ac.uk, 12 Oct 2026
Other integrations: The package can use ramcmc for robust adaptive Metropolis adaptation, and its chain output can be passed to posterior for summaries and convergence diagnostics. · github-pages.ucl.ac.uk, 12 Oct 2026
Output: Sampling returns the final chain state, main-iteration traces, and transition statistics; warm-up traces and statistics are available when trace_warm_up is enabled. · github-pages.ucl.ac.uk, 5 Oct 2026
Proposals: It supports Metropolis-Hastings transitions with random-walk, Langevin, Hamiltonian, and Barker proposals. · github-pages.ucl.ac.uk, 12 Oct 2026
Purpose: rmcmc is an R package for simulating Markov chains to compute MCMC estimates for target distributions on real-valued vector spaces. · github-pages.ucl.ac.uk, 12 Oct 2026
Sampling methods: It supports Metropolis-Hastings transitions with random-walk, Langevin, Hamiltonian, and Barker proposals; Barker is the default. · github-pages.ucl.ac.uk, 5 Oct 2026
Stan integration: The package can use Stan models through BridgeStan’s R interface when that interface is installed. · github-pages.ucl.ac.uk, 5 Oct 2026
Stan workflow: The documentation includes an article on interfacing with Stan models via BridgeStan. · github-pages.ucl.ac.uk, 12 Oct 2026
Support: The site links to GitHub to report bugs and browse the source code. · github-pages.ucl.ac.uk, 12 Oct 2026
Target distributions: The sampling function accepts a log-density formula, a list of density and gradient functions, or a BridgeStan StanModel when BridgeStan is installed. · github-pages.ucl.ac.uk, 5 Oct 2026
Warm-up adaptation: During warm-up, adapters can tune proposal scale and shape. · github-pages.ucl.ac.uk, 5 Oct 2026

Where rmcmc stands on Freedom251

Statistical Analysis Software#89 of 106