Implements methods to automate the Auer-Gervini graphical
Bayesian approach for determining the number of significant
principal components. Automation uses clustering, change points, or
simple statistical models to distinguish "long" from "short" steps
in a graph showing the posterior number of components as a function
of a prior parameter. See
Dear Emacs, please make this -*-Text-*- mode!
**************************************************
* *
* 1.1 SERIES NEWS *
* *
**************************************************
CHANGES IN R VERSION 1.1.9
UPDATES
o Modified use of S4 classes to avoid deprecated and discouraged
elements.
CHANGES IN R VERSION 1.1.0
NEW FEATURES
o Added agDimLeap, a new algorithm to find long steps.
o Added a new function, compareAgDimMethods, to report the
different dimensions reported by different algorithms.
BUG FIXES
o Fixed an error in aDimKeans and agDoimKmeans3 that only
flagged half of the "long" steps.
**************************************************
* *
* 1.0 SERIES NEWS *
* *
**************************************************
CHANGES IN R VERSION 1.0.0
NEW FEATURES
o Initial version. Class Auer-Gervini.
BUG FIXES
o Numerous.