Contains most of the popular internal and external cluster validation methods ready to use for the most of the outputs produced by functions coming from package "cluster". Package contains also functions and examples of usage for cluster stability approach that might be applied to algorithms implemented in "cluster" package as well as user defined clustering algorithms.
README:
This directory contains code, help, examples and demo for cluster validation
alghorithms working as an extention for package "cluster"
This is the third verions of package "clv". Package is believed to be tested.
Any problems, bugs, comments, documentation misunderstandings etc. please send
on the following email address: <wookashn at gmail.com>.
RELASE NOTES:
1. Small documentation fix ater CRAN check.
2. Fixing small bug in intracluster average diameter
(the result was twice smaller than should be).
TODO:
Next cluster stability approach (based on prediction) is ongoing. The code is there but not
available for end user yet. First has to be refactored and tested.
Create and document R functions for group of matrix cluster scatter measures based
indicies such as: Calinski-Harabasz index, Krzanowski & Lai index, Hartigan statistic.
Create and document R functions that helps to realise relative criteria approach for
cluster validation.
SPECIAL THANKS TO:
Artur Suchwalko for idea of this package and support.
Przemyslaw Bernacki for preparing and running manual and automated tests.
Marcin for finding a bug in intracluster average diameter index.