camir-aes2014: toolboxes/FullBNT-1.0.7/bnt/general/fgraph_to

first hg version after svn

rev	line source
wolffd@0	1 function bnet = fgraph_to_bnet(fg)
wolffd@0	2 % FGRAPH_TO_BNET Convert a factor graph to a Bayes net
wolffd@0	3 % bnet = fgraph_to_bnet(fg)
wolffd@0	4 %
wolffd@0	5 % We assume all factors are tabular_CPD.
wolffd@0	6 % We create 1 dummy observed node for every factor.
wolffd@0	7
wolffd@0	8 N = fg.nvars + fg.nfactors;
wolffd@0	9 vnodes = 1:fg.nvars;
wolffd@0	10 fnodes = fg.nvars+1:N;
wolffd@0	11 dag = zeros(N);
wolffd@0	12 for x=1:fg.nvars
wolffd@0	13 dag(x, fnodes(fg.dep{x})) = 1;
wolffd@0	14 end
wolffd@0	15 ns = [fg.node_sizes ones(1, fg.nfactors)];
wolffd@0	16 discrete = [fg.dnodes fnodes];
wolffd@0	17 bnet = mk_bnet(dag, ns, 'discrete', discrete);
wolffd@0	18 for x=1:fg.nvars
wolffd@0	19 bnet.CPD{x} = tabular_CPD(bnet, x, 'CPT', 'unif');
wolffd@0	20 end
wolffd@0	21 ev = cell(1, fg.nvars); % no evidence
wolffd@0	22 for i=1:fg.nfactors
wolffd@0	23 f = fnodes(i);
wolffd@0	24 e = fg.equiv_class(i);
wolffd@0	25 pot = convert_to_pot(fg.factors{e}, 'd', fg.dom{i}, ev);
wolffd@0	26 m = pot_to_marginal(pot);
wolffd@0	27 bnet.CPD{f} = tabular_CPD(bnet, f, 'CPT', m.T);
wolffd@0	28 end
wolffd@0	29
wolffd@0	30

Mercurial > hg > camir-aes2014