picasso: Pathwise Calibrated Sparse Shooting Algorithm

Computationally efficient tools for fitting generalized linear model with convex or non-convex penalty. Users can enjoy the superior statistical property of non-convex penalty such as SCAD and MCP which has significantly less estimation error and overfitting compared to convex penalty such as lasso and ridge. Computation is handled by multi-stage convex relaxation and the PathwIse CAlibrated Sparse Shooting algOrithm (PICASSO) which exploits warm start initialization, active set updating, and strong rule for coordinate preselection to boost computation, and attains a linear convergence to a unique sparse local optimum with optimal statistical properties. The computation is memory-optimized using the sparse matrix output.

Version: 0.5.4
Depends: R (≥ 2.15.0), lattice, igraph, MASS, Matrix
Imports: methods
Published: 2016-10-06
Author: Jason Ge, Xingguo Li, Mengdi Wang, Tong Zhang, Han Liu and Tuo Zhao
Maintainer: Jason Ge <jiange at princeton.edu>
License: GPL-2
NeedsCompilation: yes
CRAN checks: picasso results


Reference manual: picasso.pdf
Vignettes: vignette
Package source: picasso_0.5.4.tar.gz
Windows binaries: r-devel: picasso_0.5.4.zip, r-release: picasso_0.5.4.zip, r-oldrel: picasso_0.5.4.zip
OS X Mavericks binaries: r-release: picasso_0.5.4.tgz, r-oldrel: picasso_0.5.4.tgz
Old sources: picasso archive

Reverse dependencies:

Reverse imports: sparsevar


Please use the canonical form https://CRAN.R-project.org/package=picasso to link to this page.