BRBVS: An R Package for Bivariate Variable selection in Copula Survival Model(s) domain

TL;DR

BRBVS is an R package that offers a novel variable selection algorithm for bivariate survival copula models, facilitating analysis of censored survival data with user-friendly tools and visualizations.

Contribution

It introduces the first variable selection algorithm specifically designed for bivariate survival copula models, expanding the capabilities of survival analysis in R.

Findings

Provides an accessible R package for bivariate survival variable selection

Demonstrates effectiveness on censored survival data

Extensible to other model classes

Abstract

BRBVS is a publicly available \texttt{R} package on CRAN that implements the algorithm proposed in Petti et al.(2024a). The algorithm was developed as the first proposal of variable selection for the class of Bivariate Survival Copula Models originally proposed in Marra & Radice (2020) and implemented in the \texttt{GJRM} package. The core of the \texttt{BRBVS} package is to implement and make available to practitioners variable selection algorithms for bivariate survival data affected by censoring, providing easy-to-use functions and graphical outputs. The idea behind the algorithm is almost general and may also be extended to different class of models.