IMPROVED FAST GAUSS TRANSFORM Code

This program is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; version 2.1 or later. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. The author may be contacted via email at: vikas (at) umiacs (.) umd (.) edu and cyang (at) sarnoff (.) com.

I have provided separate programs for parameter selection, space subdivision, and the core IFGT algorithm. This is because when using the IFGT multiple times with differrent q the parameter selection and k-center clustering have to be done only once.

List of Core Functions
ImprovedFastGaussTransformChooseParameters	Returns the number of clusters K, the maximum truncation number p_max, and the cutoff radius r.
ImprovedFastGaussTransformChooseTruncationNumber	Given the cluster radius, returns the required truncation number p_max.
KCenterClustering	Implementation of the farthest point clustering algorithm. This is the O(N logK) version.
ImprovedFastGaussTransform	Given the space subdivison results and the parameters computes the IFGT.
DataAdaptiveImprovedFastGaussTransform	Version of the IFGT where the truncation number for each source point is different.
IFGT	A wrapper function which combines all the above. Just provide the accuracy epsilon.
Utilities
GaussTransform	Direct implementation.
example	Example program to illustrate the use of the IFGT.
generate_multiple_gaussians	Generates N samples in d dimensions from G gaussians
plot_clusters	Pretty plot of the results of the clustering procedure. Plots only for two and three dimensions.