No subject

[Jorge Cadima]

A new  version (0.2) of package 'subselect' has been uploaded to CRAN.

Package 'subselect' provides functions to carry out local search
algorithms for k-variable subsets that are optimal, as surrogates for
the full (p-variable, p > k) data sets, under three different criteria.

The functions 'rm.coef', 'rv.coef' and 'gcd.coef' assess the quality
of given k-subsets of variables under the following three 
criteria: RM (McCabe's second criterion for Principal Variables, which
measures the quality of the subset as a linear predictor of all the
variables), RV (based on Escoufier's RV-criterion, which measures the
similarity of the n point configurations defined by the original p variables
and the n point configuration resulting from the regression of all variables
on a k-subset of the variables) and GCD (based on Yanai's GCD, and which
measures the similarity of the subspaces spanned by a k-variable subset, and
a subset of g Principal Components of the full data set). 

Three additional functions, 'anneal', 'genetic' and 'improve', search for
optimal k-variable subsets under those criteria, using three different
algorithms: a simulated annealing algorithm, a genetic algorithm and a
restricted local improvement algorithm. For computational efficiency, the
basic code implementing these algorithms is written in Fortran (with calls
to Lapack subroutines), but the R functions provide a user-friendly
interface where many parameters associated with the algorithms can be
specified. Among the options, the user can control the number of
iterations, initial temperature, cooling factors  
and cooling frequency in simulated annealing, and the number of generations,
population size, admissibility of clones and presence and frequency of
mutations in the 
genetic algorithm. For all algorithms, it is possible to specify the number
of solutions required in one or more cardinalities, to specify the initial
solutions and to force the solutions to include and/or to exclude given
subsets of variables. The results of these functions are lists of R objects
detailing the resulting subsets and values of criteria, which can be
manipulated in R. 

There is a CHANGELOG file in subdirectory 'inst' documenting changes
from Version 0.1.

Here is the DESCRIPTION file for the package:

Package: subselect
Version: 0.2-0
Date: 2002/06/20
Title: Selecting variable subsets. 
Author: Jorge Orestes Cerdeira <orestes at isa.utl.pt>
        Jorge Cadima <jcadima at isa.utl.pt>
        Manuel Minhoto <minhoto at uevora.pt>
Maintainer: Jorge Cadima <jcadima at isa.utl.pt>
Description: A collection of functions which assess the quality of
variable subsets as surrogates for a full data set using three
different criteria (Yanais' GCD, Escoufier's RV and McCabe's RM), and
search for subsets which are optimal under those criteria, using a
simulated annealing algorithm, a genetic algorithm, or a restricted
local improvement algorithm.
License: GPL

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-announce mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-announce-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._