wolffd@0: if 0
wolffd@0: % Generate some sample paths
wolffd@0: 
wolffd@0: bnet = mk_map_hhmm('p', 1);
wolffd@0: % assign numbers to the nodes in topological order
wolffd@0: U = 1; A = 2; C = 3; F = 4; O = 5;
wolffd@0: 
wolffd@0: 
wolffd@0: seed = 0;
wolffd@0: rand('state', seed);
wolffd@0: randn('state', seed);
wolffd@0: 
wolffd@0: % control policy = sweep right then left
wolffd@0: T = 10;
wolffd@0: ss = 5;
wolffd@0: ev = cell(ss, T);
wolffd@0: ev(U,:) = num2cell([R*ones(1,5) L*ones(1,5)]);
wolffd@0: 
wolffd@0: % fix initial conditions to be in left most state
wolffd@0: ev{A,1} = 1; 
wolffd@0: ev{C,1} = 1; 
wolffd@0: evidence = sample_dbn(bnet, 'length', T, 'evidence', ev)
wolffd@0: 
wolffd@0: 
wolffd@0: % Now do same but with noisy actuators
wolffd@0: 
wolffd@0: bnet = mk_map_hhmm('p', 0.8);
wolffd@0: evidence = sample_dbn(bnet, 'length', T, 'evidence', ev)
wolffd@0: 
wolffd@0: end
wolffd@0: 
wolffd@0: % Now do same but with 4 observations per slice
wolffd@0: 
wolffd@0: bnet = mk_map_hhmm('p', 0.8, 'obs_model', 'four');
wolffd@0: ss = bnet.nnodes_per_slice;
wolffd@0: 
wolffd@0: ev = cell(ss, T);
wolffd@0: ev(U,:) = num2cell([R*ones(1,5) L*ones(1,5)]);
wolffd@0: ev{A,1} = 1; 
wolffd@0: ev{C,1} = 1; 
wolffd@0: evidence = sample_dbn(bnet, 'length', T, 'evidence', ev)