rstan: R Interface to Stan

User-facing R functions are provided by this package to parse, compile, test, estimate, and analyze Stan models by accessing the header-only Stan library provided by the 'StanHeaders' package. The Stan project develops a probabilistic programming language that implements full Bayesian statistical inference via Markov Chain Monte Carlo and (optionally penalized) maximum likelihood estimation via optimization. In both cases, automatic differentiation is used to quickly and accurately evaluate gradients without burdening the user with the need to derive the partial derivatives.

Version: 2.7.0-1
Depends: R (≥ 3.0.2), Rcpp (≥ 0.11.0), utils, inline, methods
Imports: stats4
LinkingTo: Rcpp, RcppEigen, BH (≥ 1.58), StanHeaders (≥ 2.7.0)
Suggests: RUnit, RcppEigen, BH (≥ 1.58), StanHeaders (≥ 2.7.0), parallel, KernSmooth, RCurl, loo
Published: 2015-07-18
Author: Jiqiang Guo [aut], Daniel Lee [ctb], Ben Goodrich [cre, aut], Joel de Guzman [ctb], Eric Niebler [ctb], Thomas Heller [ctb], John Fletcher [ctb]
Maintainer: Ben Goodrich <benjamin.goodrich at>
License: GPL (≥ 3)
NeedsCompilation: yes
Citation: rstan citation info
Materials: NEWS
CRAN checks: rstan results


Reference manual: rstan.pdf
Vignettes: RStan
Package source: rstan_2.7.0-1.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X Snow Leopard binaries: r-release: rstan_2.7.0-1.tgz, r-oldrel: not available
OS X Mavericks binaries: r-release: rstan_2.7.0-1.tgz

Reverse dependencies:

Reverse depends: brms, MIXFIM, varian
Reverse suggests: shinystan