wolffd@0: function [net, options, varargout] = netopt_weighted(net, options, x, t, eso_w, alg);
wolffd@0: %NETOPT	Optimize the weights in a network model. 
wolffd@0: %
wolffd@0: %	Description
wolffd@0: %
wolffd@0: %	NETOPT is a helper function which facilitates the training of
wolffd@0: %	networks using the general purpose optimizers as well as sampling
wolffd@0: %	from the posterior distribution of parameters using general purpose
wolffd@0: %	Markov chain Monte Carlo sampling algorithms. It can be used with any
wolffd@0: %	function that searches in parameter space using error and gradient
wolffd@0: %	functions.
wolffd@0: %
wolffd@0: %	[NET, OPTIONS] = NETOPT(NET, OPTIONS, X, T, ALG) takes a network
wolffd@0: %	data structure NET, together with a vector OPTIONS of parameters
wolffd@0: %	governing the behaviour of the optimization algorithm, a matrix X of
wolffd@0: %	input vectors and a matrix T of target vectors, and returns the
wolffd@0: %	trained network as well as an updated OPTIONS vector. The string ALG
wolffd@0: %	determines which optimization algorithm (CONJGRAD, QUASINEW, SCG,
wolffd@0: %	etc.) or Monte Carlo algorithm (such as HMC) will be used.
wolffd@0: %
wolffd@0: %	[NET, OPTIONS, VARARGOUT] = NETOPT(NET, OPTIONS, X, T, ALG) also
wolffd@0: %	returns any additional return values from the optimisation algorithm.
wolffd@0: %
wolffd@0: %	See also
wolffd@0: %	NETGRAD, BFGS, CONJGRAD, GRADDESC, HMC, SCG
wolffd@0: %
wolffd@0: 
wolffd@0: %	Copyright (c) Ian T Nabney (1996-9)
wolffd@0: 
wolffd@0: optstring = [alg, '(''neterr_weighted'', w, options, ''netgrad_weighted'', net, x, t, eso_w)'];
wolffd@0: 
wolffd@0: % Extract weights from network as single vector
wolffd@0: w = netpak(net);
wolffd@0: 
wolffd@0: % Carry out optimisation
wolffd@0: [s{1:nargout}] = eval(optstring);
wolffd@0: w = s{1};
wolffd@0: 
wolffd@0: if nargout > 1
wolffd@0:   options = s{2};
wolffd@0: 
wolffd@0:   % If there are additional arguments, extract them
wolffd@0:   nextra = nargout - 2;
wolffd@0:   if nextra > 0
wolffd@0:     for i = 1:nextra
wolffd@0:       varargout{i} = s{i+2};
wolffd@0:     end
wolffd@0:   end
wolffd@0: end
wolffd@0: 
wolffd@0: % Pack the weights back into the network
wolffd@0: net = netunpak(net, w);