loo: Efficient Leave-One-Out Cross-Validation and WAIC for Bayesian Models

Efficient approximate leave-one-out cross-validation (LOO) for Bayesian models fit using Markov chain Monte Carlo. The approximation uses Pareto smoothed importance sampling (PSIS), a new procedure for regularizing importance weights. As a byproduct of the calculations, we also obtain approximate standard errors for estimated predictive errors and for the comparison of predictive errors between models. The package also provides methods for using stacking and other model weighting techniques to average Bayesian predictive distributions.

Version: 2.0.0
Depends: R (≥ 3.1.2)
Imports: graphics, matrixStats (≥ 0.52), parallel, stats
Suggests: bayesplot (≥ 1.5.0), knitr, rmarkdown, rstan, rstanarm, rstantools, testthat
Published: 2018-04-11
Author: Aki Vehtari [aut], Andrew Gelman [aut], Jonah Gabry [cre, aut], Yuling Yao [aut], Juho Piironen [ctb], Ben Goodrich [ctb]
Maintainer: Jonah Gabry <jsg2201 at columbia.edu>
BugReports: https://github.com/stan-dev/loo/issues
License: GPL (≥ 3)
URL: http://mc-stan.org, http://discourse.mc-stan.org
NeedsCompilation: no
Citation: loo citation info
Materials: NEWS
CRAN checks: loo results

Downloads:

Reference manual: loo.pdf
Vignettes: Using the loo package
Bayesian Stacking and Pseudo-BMA weights
Writing Stan programs for use with the loo package
Package source: loo_2.0.0.tar.gz
Windows binaries: r-devel: loo_2.0.0.zip, r-release: loo_2.0.0.zip, r-oldrel: loo_2.0.0.zip
OS X binaries: r-release: loo_2.0.0.tgz, r-oldrel: loo_2.0.0.tgz
Old sources: loo archive

Reverse dependencies:

Reverse depends: evidence
Reverse imports: BAMBI, beanz, blavaan, brms, ggdmc, glmmfields, hBayesDM, MixSIAR, projpred, psycho, rstanarm, rstansim
Reverse suggests: bayesplot, CopulaDTA, idealstan, rstan, rstantools, sjstats

Linking:

Please use the canonical form https://CRAN.R-project.org/package=loo to link to this page.