wolffd@0: function CPD = bayes_update_params(CPD, self_ev, pev)
wolffd@0: % UPDATE_PARAMS_COMPLETE Bayesian parameter updating given completely observed data (tabular)
wolffd@0: % CPD = update_params_complete(CPD, self_ev, pev)
wolffd@0: %
wolffd@0: % self_ev(m) is the evidence on this node in case m.
wolffd@0: % pev(i,m) is the evidence on the i'th parent in case m (if there are any parents).
wolffd@0: % These can be arrays or cell arrays.
wolffd@0: %
wolffd@0: % We update the Dirichlet pseudo counts and set the CPT to the mean of the posterior.
wolffd@0: 
wolffd@0: if iscell(self_ev), usecell = 1; else usecell = 0; end
wolffd@0: 
wolffd@0: ncases = length(self_ev);
wolffd@0: sz = CPD.sizes;
wolffd@0: nparents = length(sz)-1;
wolffd@0: assert(nparents == size(pev,1));
wolffd@0: 
wolffd@0: if ncases == 0 | ~adjustable_CPD(CPD)
wolffd@0:   return;
wolffd@0: elseif ncases == 1 % speedup the sequential learning case by avoiding normalization of the whole array
wolffd@0:   if usecell
wolffd@0:     x = cat(1, pev{:})';
wolffd@0:     y = self_ev{1};
wolffd@0:   else
wolffd@0:     x = pev(:)';
wolffd@0:     y = self_ev;
wolffd@0:   end
wolffd@0:   switch nparents
wolffd@0:    case 0,
wolffd@0:     CPD.dirichlet(y) = CPD.dirichlet(y)+1;
wolffd@0:     CPD.CPT = CPD.dirichlet / sum(CPD.dirichlet);
wolffd@0:    case 1,
wolffd@0:     CPD.dirichlet(x(1), y) = CPD.dirichlet(x(1), y)+1;
wolffd@0:     CPD.CPT(x(1), :) = CPD.dirichlet(x(1), :) ./ sum(CPD.dirichlet(x(1), :));
wolffd@0:    case 2,
wolffd@0:     CPD.dirichlet(x(1), x(2), y) = CPD.dirichlet(x(1), x(2), y)+1;
wolffd@0:     CPD.CPT(x(1), x(2), :) = CPD.dirichlet(x(1), x(2), :) ./ sum(CPD.dirichlet(x(1), x(2), :));
wolffd@0:    case 3,
wolffd@0:     CPD.dirichlet(x(1), x(2), x(3), y) = CPD.dirichlet(x(1), x(2), x(3), y)+1;
wolffd@0:     CPD.CPT(x(1), x(2), x(3), :) = CPD.dirichlet(x(1), x(2), x(3), :) ./ sum(CPD.dirichlet(x(1), x(2), x(3), :));
wolffd@0:    otherwise,
wolffd@0:     ind = subv2ind(sz, [x y]);
wolffd@0:     CPD.dirichlet(ind) = CPD.dirichlet(ind) + 1;
wolffd@0:     CPD.CPT = mk_stochastic(CPD.dirichlet);
wolffd@0:   end
wolffd@0: else  
wolffd@0:   if usecell
wolffd@0:     data = [cell2num(pev); cell2num(self_ev)]; 
wolffd@0:   else
wolffd@0:     data = [pev; self_ev];
wolffd@0:   end
wolffd@0:   counts = compute_counts(data, sz);
wolffd@0:   CPD.dirichlet = CPD.dirichlet + counts;
wolffd@0:   CPD.CPT = mk_stochastic(CPD.dirichlet);
wolffd@0: end