wolffd@0: function [d2, win_nodes] = somfwd(net, x)
wolffd@0: %SOMFWD	Forward propagation through a Self-Organising Map.
wolffd@0: %
wolffd@0: %	Description
wolffd@0: %	D2 = SOMFWD(NET, X) propagates the data matrix X through  a SOM NET,
wolffd@0: %	returning the squared distance matrix D2 with dimension NIN by
wolffd@0: %	NUM_NODES.  The $i$th row represents the squared Euclidean distance
wolffd@0: %	to each of the nodes of the SOM.
wolffd@0: %
wolffd@0: %	[D2, WIN_NODES] = SOMFWD(NET, X) also returns the indices of the
wolffd@0: %	winning nodes for each pattern.
wolffd@0: %
wolffd@0: %	See also
wolffd@0: %	SOM, SOMTRAIN
wolffd@0: %
wolffd@0: 
wolffd@0: %	Copyright (c) Ian T Nabney (1996-2001)
wolffd@0: 
wolffd@0: % Check for consistency
wolffd@0: errstring = consist(net, 'som', x);
wolffd@0: if ~isempty(errstring)
wolffd@0:     error(errstring);
wolffd@0: end
wolffd@0: 
wolffd@0: % Turn nodes into matrix of centres
wolffd@0: nodes = (reshape(net.map, net.nin, net.num_nodes))';
wolffd@0: % Compute squared distance matrix
wolffd@0: d2 = dist2(x, nodes);
wolffd@0: % Find winning node for each pattern: minimum value in each row
wolffd@0: [w, win_nodes] = min(d2, [], 2);