isotone: Active set and generalized PAVA for isotone optimization

Contains two main functions: gpava() for solving general isotone regression problem using the pool-adjacent-violators algorithm (PAVA). activeSet() provides a framework of active set methods for isotone optimization problems with arbitrary order restrictions. Various types of loss functions are pre-specified.

Version: 1.0-2
Published: 2014-09-10
Author: Patrick Mair [aut, cre], Jan De Leeuw [aut], Kurt Hornik [aut]
Maintainer: Patrick Mair <mair at>
License: GPL-2
Copyright: see file COPYRIGHTS
NeedsCompilation: yes
Citation: isotone citation info
CRAN checks: isotone results


Reference manual: isotone.pdf
Package source: isotone_1.0-2.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X Snow Leopard binaries: r-release: isotone_1.0-2.tgz, r-oldrel: isotone_1.0-2.tgz
OS X Mavericks binaries: r-release: isotone_1.0-2.tgz
Old sources: isotone archive

Reverse dependencies:

Reverse depends: gpmap, twostageTE
Reverse imports: comparison
Reverse suggests: RDS, WeightedCluster