OSCARS: Global Bounded Optimization by the OSCARS-II Algorithm

A collection of general optimization routines based on variants of the One Side Cut Accelerated Random Search (OSCARS-II) algorithm (Price et al., 2020, <doi:10.1007/s10898-020-00928-6>). The main function , 'oscars()', performs black-box optimization of a general (including nonsmooth or discontinuous) function subject to simple bounds on the unknowns. If all bounds are finite, oscars searches globally. The main method implements a stochastic direct search method and is derivative free. Testing shows the OSCARS-II algorithm usually finds extrema with fewer function evaluations than similar global derivative-free methods.

Version:	0.1.2
Imports:	stats
Published:	2026-05-06
DOI:	10.32614/CRAN.package.OSCARS
Author:	Chris Price [aut, cre], Trent McDonald [aut, ctb] (R packaging)
Maintainer:	Chris Price <chrisj.price at canterbury.ac.nz>
License:	Apache License (≥ 2)
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	OSCARS results

Documentation:

Reference manual:

OSCARS.html , OSCARS.pdf

Downloads:

Package source:	OSCARS_0.1.2.tar.gz
Windows binaries:	r-devel: OSCARS_0.1.2.zip, r-release: not available, r-oldrel: not available
macOS binaries:	r-release (arm64): OSCARS_0.1.2.tgz, r-oldrel (arm64): OSCARS_0.1.2.tgz, r-release (x86_64): OSCARS_0.1.2.tgz, r-oldrel (x86_64): OSCARS_0.1.2.tgz

Linking:

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