IMPROVED FAST GAUSS TRANSFORM
Copyright Information
The code was written by Vikas C. Raykar and Changjiang Yang is copyrighted under the Lesser GPL:
Copyright (C) 2006 Vikas C. Raykar and Changjiang Yang
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. |
10/18/2006