[R] [R-pkgs] Release of optimbase, optimsimplex and neldermead packages

[Sebastien Bihorel]

Dear R users,

I am pleased to announce the release of three new R packages: optimbase,
optimsimplex, and neldermead.
- optimbase provides a set of commands to manage an abstract optimization
method. The goal is to provide a building block for a large class of
specialized optimization methods. This package manages: the number of
variables, the minimum and maximum bounds, the number of non linear
inequality constraints, the cost function, the logging system, various
termination criteria, etc...
- optimsimplex provides a building block for optimization algorithms based
on a simplex. The optimsimplex package may be used in the following
optimization methods: the simplex method of Spendley et al., the method of
Nelder and Mead, Box's algorithm for constrained optimization, the
multi-dimensional search by Torczon, etc...
- neldermead depends on optimbase and optimsimplex and 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 size simplex
method of Spendley, Hext and Himsworth (unconstrained optimization with a
fixed sized simplex), the variable size simplex method of Nelder and Mead
(unconstrained optimization with a variable sized simplex), Box's complex
method (constrained optimization with a variable sized simplex). This
package includes an R-port of the fminsearch function (available in Matlab
and Scilab) which is a specialized use of the more general neldermead
package and computes the unconstrained minimimum of given function with the
Nelder-Mead algorithm.

One important benefit offered by those packages (especially optimbase) is
the possibly for the user to define functions that will be called at the
each iteration of the optimization process. These functions may be used when
debugging a specialized optimization algorithm, to write to one or several
report files, or create/update optimization graphs. optimbase can
accommodate both derivative-based or derivative-free algorithms.

The three packages are ports of original Scilab modules written by Michael
Baudin at the Digiteo Consortium (and previously at Institut National de
Recheche en Informatique et en Automatique).

All packages are available in version 1.0-1 on CRAN.

Any question, comment or feedback on those packages can be sent at:
sb.pmlab at gmail.com.

Sebastien Bihorel

	[[alternative HTML version deleted]]

_______________________________________________
R-packages mailing list
R-packages at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-packages