crossfit: A Graph-Based Cross-Fitting Engine in R

Provides a general cross-fitting engine for semiparametric estimation (e.g., double/debiased machine learning). Supports user-defined target functionals and directed acyclic graphs of nuisance learners with per-node training fold widths, target-specific evaluation windows, and fold-allocation modes ("overlap", "disjoint", "independence"). Returns either numeric estimates (mode = "estimate") or cross-fitted prediction functions (mode = "predict"), with configurable aggregation over panels and repetitions, reuse-aware caching, and failure isolation, making it well-suited for simulation studies and large benchmarks.

Version:	0.1.3
Depends:	R (≥ 4.1.0)
Imports:	stats, utils
Suggests:	knitr, rmarkdown, testthat (≥ 3.0.0)
Published:	2026-03-04
DOI:	10.32614/CRAN.package.crossfit
Author:	Etienne Peyrot [aut, cre]
Maintainer:	Etienne Peyrot <etienne.peyrot at inserm.fr>
BugReports:	https://github.com/EtiennePeyrot/crossfit-R/issues
License:	GPL-3
URL:	https://github.com/EtiennePeyrot/crossfit-R
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	crossfit results

Documentation:

Reference manual:	crossfit.html , crossfit.pdf
Vignettes:	Introduction to crossfit (source, R code)

Downloads:

Package source:	crossfit_0.1.3.tar.gz
Windows binaries:	r-devel: crossfit_0.1.3.zip, r-release: crossfit_0.1.3.zip, r-oldrel: crossfit_0.1.3.zip
macOS binaries:	r-release (arm64): crossfit_0.1.3.tgz, r-oldrel (arm64): crossfit_0.1.3.tgz, r-release (x86_64): crossfit_0.1.3.tgz, r-oldrel (x86_64): crossfit_0.1.3.tgz
Old sources:	crossfit archive

Linking:

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