optiscale: Optimal Scaling

Optimal scaling of a data vector is obtained through a least-squares transformation subject to appropriate measurement constraints. Further information about this methodology is provided by Young (1981)<doi:10.1007/BF02293796> and Jacoby (1999)< doi:10.2307/2991794>.

Version: 1.2
Depends: lattice
Published: 2020-02-28
Author: William G. Jacoby
Maintainer: William G. Jacoby <wm.g.jacoby at gmail.com>
License: GPL-2
NeedsCompilation: no
In views: Psychometrics
CRAN checks: optiscale results


Reference manual: optiscale.pdf
Package source: optiscale_1.2.tar.gz
Windows binaries: r-devel: optiscale_1.2.zip, r-release: optiscale_1.2.zip, r-oldrel: optiscale_1.2.zip
macOS binaries: r-release: optiscale_1.2.tgz, r-oldrel: optiscale_1.2.tgz
Old sources: optiscale archive

Reverse dependencies:

Reverse imports: DAMisc


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