imv: Model Comparison via the 'InterModel Vigorish' ('IMV')

Computes the 'InterModel Vigorish' ('IMV'), a metric for comparing the predictive accuracy of two models for binary outcomes. The 'IMV' is derived from the expected value of a bettor using one model's predicted probabilities against those of a competing model, and is estimated via k-fold cross-validation. Methods are provided for generalized linear models, mixed-effects models ('lme4'), and item response theory models ('mirt'). See <doi:10.1371/journal.pone.0316491>.

Version:	0.3
Depends:	R (≥ 3.5.0)
Suggests:	lme4, mirt, testthat (≥ 3.0.0)
Published:	2026-05-11
DOI:	10.32614/CRAN.package.imv
Author:	Ben Domingue [aut, cre], Christian Jackson [ctb]
Maintainer:	Ben Domingue <ben.domingue at gmail.com>
License:	MIT + file LICENSE
NeedsCompilation:	no
Language:	en-US
Materials:	README, NEWS
CRAN checks:	imv results

Documentation:

Reference manual:

imv.html , imv.pdf

Downloads:

Package source:	imv_0.3.tar.gz
Windows binaries:	r-devel: imv_0.3.zip, r-release: imv_0.3.zip, r-oldrel: imv_0.3.zip
macOS binaries:	r-release (arm64): imv_0.3.tgz, r-oldrel (arm64): imv_0.3.tgz, r-release (x86_64): imv_0.3.tgz, r-oldrel (x86_64): imv_0.3.tgz

Linking:

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