r-loo
Leave-One-Out cross-validation and WAIC for Bayesian models
This package provides an implementation of efficient approximate leave-one-out (LOO) cross-validation for Bayesian models fit using Markov chain Monte Carlo, as described in doi:10.1007/s11222-016-9696-4. 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.
- Versions: 2.8.0
- Website: https://mc-stan.org/loo/
- Licenses: GPL 3+
- Package source: gnu/packages/cran.scm
- Builds: See build status
- Issues: See known issues
Installation
Install the latest version of r-loo as follows:
guix install r-loo
Or install a particular version:
guix install r-loo@2.8.0
You can also install packages in augmented, pure or containerized environments for development or simply to try them out without polluting your user profile. See the guix shell documentation for more information.
Badge code
You can use the following badge to inform users of r-loo about the latest version available in Guix.
Example HTML:
<a href='https://packages.guix.gnu.org/packages/r-loo'><img src='https://packages.guix.gnu.org/packages/r-loo/badges/latest-version.svg' alt='Version in GNU Guix'></a>
Example Markdown:
[](https://packages.guix.gnu.org/packages/r-loo)
Example Org:
[[https://packages.guix.gnu.org/packages/r-loo][https://packages.guix.gnu.org/packages/r-loo/badges/latest-version.svg]]