cort: Some Empiric and Nonparametric Copula Models

Provides S4 classes and methods to fit several copula models: The classic empirical checkerboard copula and the empirical checkerboard copula with known margins, see Cuberos, Masiello and Maume-Deschamps (2019) <doi:10.1080/03610926.2019.1586936> are proposed. These two models allow to fit copulas in high dimension with a small number of observations, and they are always proper copulas. Some flexibility is added via a possibility to differentiate the checkerboard parameter by dimension. The last model consist of the implementation of the Copula Recursive Tree algorithm proposed by Laverny, Maume-Deschamps, Masiello and Rullière (2020) <arXiv:2005.02912>, including the localised dimension reduction, which fits a copula by recursive splitting of the copula domain. We also provide an efficient way of mixing copulas, allowing to bag the algorithm into a forest, and a generic way of measuring d-dimensional boxes with a copula.

Version: 0.3.2
Depends: R (≥ 2.10)
Imports: Rdpack, methods, purrr, nloptr, osqp, Rcpp, furrr (≥ 0.2.0)
LinkingTo: Rcpp
Suggests: covr, testthat (≥ 2.1.0), spelling, knitr, rmarkdown
Published: 2020-12-01
Author: Oskar Laverny ORCID iD [aut, cre]
Maintainer: Oskar Laverny <oskar.laverny at>
License: MIT + file LICENSE
NeedsCompilation: yes
Language: en-US
Materials: README NEWS
In views: Distributions
CRAN checks: cort results


Reference manual: cort.pdf
Vignettes: 1. Empirical Checkerboard Copula
2. The Copula Recursive Tree
3. Empirical Checkerboard Copula with known margins
4. Convex mixture of m-randomized checkerboards
Package source: cort_0.3.2.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
macOS binaries: r-release: cort_0.3.2.tgz, r-oldrel: cort_0.3.2.tgz
Old sources: cort archive


Please use the canonical form to link to this page.