We efficiently approximate leave-one-out cross-validation (LOO) using Pareto smoothed importance sampling (PSIS), a new procedure for regularizing importance weights. As a byproduct of our calculations, we also obtain approximate standard errors for estimated predictive errors, and for the comparison of predictive errors between two models. We also compute the widely applicable information criterion (WAIC).
| Version: | 0.1.3 |
| Depends: | R (≥ 3.1.2) |
| Imports: | graphics, matrixStats (≥ 0.14.1), parallel, stats |
| Suggests: | knitr, testthat |
| Published: | 2015-09-18 |
| Author: | Aki Vehtari [aut], Andrew Gelman [aut], Jonah Gabry [cre, aut], Juho Piironen [ctb], Ben Goodrich [ctb] |
| Maintainer: | Jonah Gabry <jsg2201 at columbia.edu> |
| BugReports: | https://github.com/jgabry/loo/issues |
| License: | GPL (≥ 3) |
| URL: | https://github.com/jgabry/loo |
| NeedsCompilation: | no |
| Citation: | loo citation info |
| CRAN checks: | loo results |
| Reference manual: | loo.pdf |
| Vignettes: |
Example |
| Package source: | loo_0.1.3.tar.gz |
| Windows binaries: | r-devel: loo_0.1.3.zip, r-release: loo_0.1.3.zip, r-oldrel: loo_0.1.3.zip |
| OS X Snow Leopard binaries: | r-release: loo_0.1.2.tgz, r-oldrel: not available |
| OS X Mavericks binaries: | r-release: loo_0.1.3.tgz |
| Old sources: | loo archive |
| Reverse imports: | blavaan, brms |
| Reverse suggests: | rstan |