neldermead: R port of the Scilab neldermead module

Provides several direct search optimization algorithms based on the simplex method. The provided algorithms are direct search algorithms, i.e. algorithms which do not use the derivative of the cost function. They are based on the update of a simplex. The following algorithms are available: the fixed shape simplex method of Spendley, Hext and Himsworth (unconstrained optimization with a fixed shape simplex), the variable shape simplex method of Nelder and Mead (unconstrained optimization with a variable shape simplex made), and Box's complex method (constrained optimization with a variable shape simplex).

Version: 1.0-9
Depends: optimbase (≥ 1.0-9), optimsimplex (≥ 1.0-5), methods
Published: 2014-03-03
Author: Sebastien Bihorel, Michael Baudin (author of the original module)
Maintainer: Sebastien Bihorel <sb.pmlab at>
License: CeCILL-2
NeedsCompilation: no
Materials: ChangeLog
In views: Optimization
CRAN checks: neldermead results


Reference manual: neldermead.pdf
Vignettes: Introduction to the neldermead package
Package source: neldermead_1.0-9.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X Snow Leopard binaries: r-release: neldermead_1.0-9.tgz, r-oldrel: neldermead_1.0-9.tgz
OS X Mavericks binaries: r-release: neldermead_1.0-9.tgz
Old sources: neldermead archive

Reverse dependencies:

Reverse depends: classyfire, scaRabee