Fast variable selection in high-dimensional settings with control of the false discovery rate (FDR).
The R package of the Terminating-Random Experiments (T-Rex) selector, with bindings to the TRexSelector C++ library. The Python package binds the same library. The vignettes walk through the package with code that runs when they are built, so every output shown is the output of the code above it, and the reference manual describes every class and function from its help page. The user guides of the Python package and the C++ library present the same content page for page.
The method family grows within a series of papers, among which the following three form the theoretical core of the framework so far:
J. Machkour, M. Muma, and D. P. Palomar, “The terminating-random experiments selector: Fast high-dimensional variable selection with false discovery rate control,” Signal Processing, vol. 231, article 109894, 2025. (https://doi.org/10.1016/j.sigpro.2025.109894)
J. Machkour, M. Muma, and D. P. Palomar, “High-dimensional false discovery rate control for dependent variables,” Signal Processing, vol. 234, article 109990, 2025. (https://doi.org/10.1016/j.sigpro.2025.109990)
T. Koka, J. Machkour, D. P. Palomar, and M. Muma, “Virtual dummies: Enabling scalable FDR-controlled variable selection via sequential sampling of null features,” arXiv:2604.07464, 2026. (https://arxiv.org/abs/2604.07464)
Every page is a vignette of the package.
browseVignettes("TRexSelector") lists them together with
this overview, and vignette("v01-getting-started") opens a
page in the viewer of the R session. The pages cite the literature, and
every page ends with its bibliography. The vignettes are built with
knitr [1]
and rmarkdown [2], and their figures are drawn
with ggplot2 [3].
help(package = "TRexSelector") lists the help pages of
the installed package, and the left pane links them when the page is
read from an installed package.citation("TRexSelector") gives the package entry and
the method papers.