ECMLE: Approximating Evidence via Bounded Harmonic Means

Implements the Elliptical Covering Marginal Likelihood Estimator (ECMLE), a geometric method for approximating marginal likelihood from posterior draws and log-posterior evaluations. The method constructs a collection of non-overlapping ellipsoids in a high-posterior-density region, computes the covered volume, and combines this with posterior sample coverage to estimate model evidence. It is designed to stabilize harmonic-mean-based evidence approximation and can be applied in multimodal settings. The methodology is described in Naderi et al. (2025) <doi:10.48550/arXiv.2510.20617>.

Version:	0.1.0
Depends:	R (≥ 3.6.0)
Imports:	IDPmisc, graphics, stats, withr
Suggests:	testthat (≥ 3.0.0), knitr, rmarkdown
Published:	2026-03-26
DOI:	10.32614/CRAN.package.ECMLE
Author:	Dana Naderi [aut, cre], Christian P. Robert [aut], Kaniav Kamary [aut], Darren Wraith [aut]
Maintainer:	Dana Naderi <naderi at ceremade.dauphine.fr>
BugReports:	https://github.com/da-na-deri/ECMLE/issues
License:	GPL-3
URL:	https://github.com/da-na-deri/ECMLE
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	ECMLE results

Documentation:

Reference manual:	ECMLE.html , ECMLE.pdf
Vignettes:	Getting started with ECMLE (source, R code)

Downloads:

Package source:	ECMLE_0.1.0.tar.gz
Windows binaries:	r-devel: ECMLE_0.1.0.zip, r-release: ECMLE_0.1.0.zip, r-oldrel: ECMLE_0.1.0.zip
macOS binaries:	r-release (arm64): ECMLE_0.1.0.tgz, r-oldrel (arm64): not available, r-release (x86_64): ECMLE_0.1.0.tgz, r-oldrel (x86_64): ECMLE_0.1.0.tgz

Linking:

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