parcor-package {parcor} | R Documentation |
This package contains different methods to estimate the matrix of partial correlations based on a (n x p) matrix X of observation. For low-dimensional settings (p>n), the matrix of partial correlations can be estimated based on p least-squares regression fits. However, in high-dimensional scenarios (p<n), theses least-squares problems are ill-posed and need to be regularized. This package contains four different regularized regression techniques for the estimation of the partial correlations: lasso, adaptive lasso, ridge regression, and Partial Least Squares. In addition, the package provides model selection for lasso, adaptive lasso and Ridge regression based on cross-validation.
Package: | parcor |
Type: | Package |
Version: | 0.2-1 |
Date: | 2009-11-17 |
License: | GPL2 or newer |
LazyLoad: | yes |
Nicole Kraemer, Juliane Schaefer
Maintainer: Nicole Kraemer <nkraemer@cs.tu-berlin.de>
N. Kraemer, J. Schaefer, A.-L. Boulesteix (2009) "Regularized Estimation of Large-Scale Gene Regulatory Networks with Gaussian Graphical Models", BMC Bioinformatics, 10:384
http://www.biomedcentral.com/1471-2105/10/384/