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