| smacof-package {smacof} | R Documentation |
This package provides the following approaches of multidimensional scaling (MDS) based on stress minimization by means of majorization (smacof): Simple smacof on symmetric dissimilarity matrices, smacof for rectangular matrices (unfolding models), smacof with constraints on the configuration, three-way smacof for individual differences (including constraints for idioscal, indscal, and identity), and spherical smacof (primal and dual algorithm). Each of these approaches is implemented in a metric and nonmetric manner including primary, secondary, and tertiary approaches for tie handling.
| Package: | smacof |
| Type: | Package |
| Version: | 0.9-0 |
| Date: | 2008-05-24 |
| License: | GPL |
The function for basic SMACOF on symmetric dissimilarity matrices is smacofSym(). For rectangular input matrices (unfolding model) smacofRect() is appropriate and by means of smacofIndDiff() individual difference models (three-way MDS) can be computed.
Jan de Leeuw, Patrick Mair
Maintainer: Jan de Leeuw <deleeuw@stat.ucla.edu>
de Leeuw, J. & Mair, P. (2008). Multidimensional scaling using majorization: The R package smacof. Journal of Statistical Software, forthcoming.
smacofSym, smacofRect, smacof, smacofIndDiff, smacofSphere.primal, smacofSphere.dual
data(trading) res <- smacofSym(trading) res