MOSAlloc: Constraint Multiobjective Sample Allocation

Provides a framework for multipurpose optimal resource allocation in survey sampling, extending the classical optimal allocation principles introduced by Tschuprow (1923) and Neyman (1934) to multidomain and multivariate allocation problems. The primary method mosalloc() allows for the consideration of precision and cost constraints at the subpopulation level while minimizing either a vector of sampling errors or survey costs across a broad range of optimal sample allocation problems. The approach supports both single- and multistage designs. For single-stage stratified random sampling, the mosallocSTRS() function offers a user- friendly interface. Sensitivity analysis is supported through the problem's dual variables, which are naturally obtained via the internal use of the Embedded Conic Solver from the 'ECOSolveR' package. See Willems (2025, <doi:10.25353/ubtr-9200-484c-5c89>) for a detailed description of the theory behind 'MOSAlloc'.

Version:	1.2.3
Imports:	ECOSolveR, Matrix
Suggests:	parallel, testthat (≥ 3.0.0)
Published:	2026-01-27
DOI:	10.32614/CRAN.package.MOSAlloc
Author:	Felix Willems [aut, cre], Ralf Münnich [ths]
Maintainer:	Felix Willems <mail.willemsf+MOSAlloc at gmail.com>
License:	GPL (≥ 3)
NeedsCompilation:	no
Citation:	MOSAlloc citation info
CRAN checks:	MOSAlloc results [issues need fixing before 2026-02-11]

Documentation:

Reference manual:

MOSAlloc.html , MOSAlloc.pdf

Downloads:

Package source:	MOSAlloc_1.2.3.tar.gz
Windows binaries:	r-devel: MOSAlloc_1.2.3.zip, r-release: MOSAlloc_1.2.3.zip, r-oldrel: MOSAlloc_1.2.3.zip
macOS binaries:	r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): MOSAlloc_1.2.3.tgz, r-oldrel (x86_64): MOSAlloc_1.2.3.tgz

Linking:

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