1- Package: loo
21Type: Package
2+ Package: loo
33Title: Efficient Leave-One-Out Cross-Validation and WAIC for Bayesian Models
44Version: 2.8.0.9000
55Date: 2024-07-03
6- Authors@R: c(person("Aki", "Vehtari", email = "Aki.Vehtari@aalto.fi", role = c("aut")),
7- person("Jonah", "Gabry", email = "jsg2201@columbia.edu", role = c("cre", "aut")),
8- person("Måns", "Magnusson", role = c("aut")),
9- person("Yuling", "Yao", role = c("aut")),
10- person("Paul-Christian", "Bürkner", role = c("aut")),
11- person("Topi", "Paananen", role = c("aut")),
12- person("Andrew", "Gelman", role = c("aut")),
13- person("Ben", "Goodrich", role = c("ctb")),
14- person("Juho", "Piironen", role = c("ctb")),
15- person("Bruno", "Nicenboim", role = c("ctb")),
16- person("Leevi", "Lindgren", role = c("ctb")))
6+ Authors@R: c(
7+ person("Aki", "Vehtari", , "Aki.Vehtari@aalto.fi", role = "aut"),
8+ person("Jonah", "Gabry", , "jsg2201@columbia.edu", role = c("cre", "aut")),
9+ person("Måns", "Magnusson", role = "aut"),
10+ person("Yuling", "Yao", role = "aut"),
11+ person("Paul-Christian", "Bürkner", role = "aut"),
12+ person("Topi", "Paananen", role = "aut"),
13+ person("Andrew", "Gelman", role = "aut"),
14+ person("Ben", "Goodrich", role = "ctb"),
15+ person("Juho", "Piironen", role = "ctb"),
16+ person("Bruno", "Nicenboim", role = "ctb"),
17+ person("Leevi", "Lindgren", role = "ctb")
18+ )
1719Maintainer: Jonah Gabry <jsg2201@columbia.edu>
18- URL: https://mc-stan.org/loo/, https://discourse.mc-stan.org
19- BugReports: https://github.com/stan-dev/loo/issues
2020Description: Efficient approximate leave-one-out cross-validation (LOO)
21- for Bayesian models fit using Markov chain Monte Carlo, as
22- described in Vehtari, Gelman, and Gabry (2017)
23- <doi:10.1007/s11222-016-9696-4>.
24- The approximation uses Pareto smoothed importance sampling (PSIS),
25- a new procedure for regularizing importance weights.
26- As a byproduct of the calculations, we also obtain approximate
27- standard errors for estimated predictive errors and for the comparison
28- of predictive errors between models. The package also provides methods
29- for using stacking and other model weighting techniques to average
30- Bayesian predictive distributions.
21+ for Bayesian models fit using Markov chain Monte Carlo, as described
22+ in Vehtari, Gelman, and Gabry (2017) <doi:10.1007/s11222-016-9696-4>.
23+ The approximation uses Pareto smoothed importance sampling (PSIS), a
24+ new procedure for regularizing importance weights. As a byproduct of
25+ the calculations, we also obtain approximate standard errors for
26+ estimated predictive errors and for the comparison of predictive
27+ errors between models. The package also provides methods for using
28+ stacking and other model weighting techniques to average Bayesian
29+ predictive distributions.
3130License: GPL (>=3)
32- LazyData: TRUE
31+ URL: https://mc-stan.org/loo/, https://discourse.mc-stan.org
32+ BugReports: https://github.com/stan-dev/loo/issues
3333Depends:
3434 R (>= 3.1.2)
3535Imports:
@@ -50,8 +50,13 @@ Suggests:
5050 rstantools,
5151 spdep,
5252 testthat (>= 2.1.0)
53- VignetteBuilder: knitr
53+ VignetteBuilder:
54+ knitr
55+ Config/testthat/edition: 3
56+ Config/testthat/parallel: true
57+ Config/testthat/start-first: loo_subsampling_cases, loo_subsampling
5458Encoding: UTF-8
55- SystemRequirements: pandoc (>= 1.12.3), pandoc-citeproc
56- RoxygenNote: 7.3.2
59+ LazyData: TRUE
5760Roxygen: list(markdown = TRUE)
61+ RoxygenNote: 7.3.2
62+ SystemRequirements: pandoc (>= 1.12.3), pandoc-citeproc
0 commit comments