caspar: Clustered and sparse regression (CaSpaR)

Clustered and sparse regression gives structured sparse solutions to regression problems. In particular, we consider cases where the structural relationship between predictors can be cast as a pairwise distance. The package gives fitting and cross validation routines for parameter selection.

Version: 1.0
Depends: R (≥ 2.10.1), mvtnorm
Suggests: mvtnorm
Published: 2010-12-04
Author: Daniel Percival
Maintainer: Daniel Percival <dperciva at andrew.cmu.edu>
License: GPL-2
CRAN checks: caspar results

Downloads:

Package source: caspar_1.0.tar.gz
MacOS X binary: caspar_1.0.tgz
Windows binary: caspar_1.0.zip
Reference manual: caspar.pdf