* Classification Topology Project *

Implements a methodology based on local neighborhood graphs (Gabriel graphs)
to characterize and to visualize the topology of the classification dataset
(distributions, overlap) and of the decision surface.

Code is developed as a set of S4 classes to compute the local neighborhood graph,
plotting to visualize the graph, graph ediding for producing condenced graphs that
highlight the decision surface topology, and summary methods for the graphs
to report a summary of the graph topology.

Additional methods for resampling the local neighborhood graphs will be added.