MMAD: Minorization-Maximization via Assembly-Decomposition Technology

A formula-driven framework for maximizing target functions via the minorization-maximization (MM) algorithm. The package represents the target as a symbolic expression tree, infers its curvature via disciplined-convex-programming rules, and constructs a separable surrogate at each iterate using only Jensen's inequality and the supporting hyperplane. The driver maximizes the surrogate via block-coordinate Newton with line search, falling back to a multivariate step on any non-separable residue. A formula interface accepts standard R expressions (including 'sum()' reductions and 'X %*% theta' design-matrix products) so statistical models such as Poisson regression can be written in one line.

Version: 3.0.0
Depends: R (≥ 2.10)
Suggests: testthat (≥ 3.0.0)
Published: 2026-07-07
DOI: 10.32614/CRAN.package.MMAD
Author: Xifen Huang [aut], Jinfeng Xu [aut], Jiaqi Gu [aut, cre]
Maintainer: Jiaqi Gu <jiaqigu at usf.edu>
License: GPL-3
NeedsCompilation: no
CRAN checks: MMAD results

Documentation:

Reference manual: MMAD.html , MMAD.pdf

Downloads:

Package source: MMAD_3.0.0.tar.gz
Windows binaries: r-devel: MMAD_3.0.0.zip, r-release: MMAD_2.0.1.zip, r-oldrel: MMAD_3.0.0.zip
macOS binaries: r-release (arm64): MMAD_3.0.0.tgz, r-oldrel (arm64): MMAD_3.0.0.tgz, r-release (x86_64): MMAD_3.0.0.tgz, r-oldrel (x86_64): MMAD_3.0.0.tgz
Old sources: MMAD archive

Linking:

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