provenance (version 0.6)

procrustes: Generalised Procrustes Analysis (GPA)

Description

Given a number of input datasets, this function performs an MDS analysis on each of these and the feeds the resulting configurations into a GPA algorithm, which uses a combination of transformations (reflections, rotations, translations and scaling) to find a 'consensus' configuration which best matches all the component configurations in a least-squares sense.

Usage

procrustes(...)

Arguments

...

a sequence of datasets of classes distributional and compositional

Value

• an object of class GPA, i.e. a list containing the following items:

  points: a two column vector with the coordinates of the group configuration

  labels: a list with the sample names

References

Dryden, Ian, and Maintainer Ian Dryden. "Shapes package." Vienna, Austria: R Foundation for Statistical Computing (2012).

Examples

data(Namib)
GPA <- procrustes(Namib$DZ,Namib$HM)
plot(GPA)