compute_nn_graph.m

toolbox_dimreduc - a toolbox for dimension reduction methods This toolbox is an educational and rec

function [D,A] = compute_nn_graph(X,options)

% compute_nn_graph - compute nearest neighbor graph
%
% [D,A] = compute_nn_graph(X,options);
%
%	X is an array of points, size must be (dimension x nbr_points)
%       X(:,i) is the ith point.
%
%   D is the graph matrix : D(i,j)==Inf if point i and j are not connected,
%       D(i,j)=|X(:,i)-X(:,j)| otherwise
%
% The graph can be computed using two modes:
%   options.nn_epsilon=dist_between_NN  ==> use epsilon-based NN mode
%   options.nn_nbr=#NN                  ==> use k-based NN mode.
%   options.nn_automatic=#NN            ==> evaluate epsilon to match the
%           k-NN (tends to prune outliers.
%
%   Copyright (c) 2005 Gabriel Peyr