Daniel@0: function dag = learn_struct_K2(data, ns, order, varargin)
Daniel@0: % LEARN_STRUCT_K2 Greedily learn the best structure compatible with a fixed node ordering
Daniel@0: % best_dag = learn_struct_K2(data, node_sizes, order, ...)
Daniel@0: %
Daniel@0: % data(i,m) = value of node i in case m (can be a cell array).
Daniel@0: % node_sizes(i) is the size of node i.
Daniel@0: % order(i) is the i'th node in the topological ordering.
Daniel@0: %
Daniel@0: % The following optional arguments can be specified in the form of name/value pairs:
Daniel@0: % [default value in brackets]
Daniel@0: %
Daniel@0: % max_fan_in - this the largest number of parents we allow per node [N]
Daniel@0: % scoring_fn - 'bayesian' or 'bic' [ 'bayesian' ]
Daniel@0: %              Currently, only networks with all tabular nodes support Bayesian scoring.
Daniel@0: % type       - type{i} is the type of CPD to use for node i, where the type is a string
Daniel@0: %              of the form 'tabular', 'noisy_or', 'gaussian', etc. [ all cells contain 'tabular' ]
Daniel@0: % params     - params{i} contains optional arguments passed to the CPD constructor for node i,
Daniel@0: %              or [] if none.  [ all cells contain {'prior', 1}, meaning use uniform Dirichlet priors ]
Daniel@0: % discrete   - the list of discrete nodes [ 1:N ]
Daniel@0: % clamped    - clamped(i,m) = 1 if node i is clamped in case m [ zeros(N, ncases) ]
Daniel@0: % verbose    - 'yes' means display output while running [ 'no' ]
Daniel@0: %
Daniel@0: % e.g., dag = learn_struct_K2(data, ns, order, 'scoring_fn', 'bic', 'params', [])
Daniel@0: %
Daniel@0: % To be backwards compatible with BNT2, you can also specify arguments as follows
Daniel@0: %   dag = learn_struct_K2(data, node_sizes, order, max_fan_in)    
Daniel@0: %
Daniel@0: % This algorithm is described in
Daniel@0: % - Cooper and Herskovits,  "A Bayesian method for the induction of probabilistic
Daniel@0: %      networks from data", Machine Learning Journal 9:308--347, 1992
Daniel@0: 
Daniel@0: [n ncases] = size(data);
Daniel@0: 
Daniel@0: % set default params
Daniel@0: type = cell(1,n);
Daniel@0: params = cell(1,n);
Daniel@0: for i=1:n
Daniel@0:   type{i} = 'tabular';
Daniel@0:   %params{i} = { 'prior', 1 };
Daniel@0:   params{i} = { 'prior_type', 'dirichlet', 'dirichlet_weight', 1 };
Daniel@0: end
Daniel@0: scoring_fn = 'bayesian';
Daniel@0: discrete = 1:n;
Daniel@0: clamped = zeros(n, ncases);
Daniel@0: 
Daniel@0: max_fan_in = n;
Daniel@0: verbose = 0;
Daniel@0: 
Daniel@0: args = varargin;
Daniel@0: nargs = length(args);
Daniel@0: if length(args) > 0 
Daniel@0:   if isstr(args{1})
Daniel@0:     for i=1:2:nargs
Daniel@0:       switch args{i},
Daniel@0:        case 'verbose',    verbose = strcmp(args{i+1}, 'yes');
Daniel@0:        case 'max_fan_in', max_fan_in = args{i+1}; 
Daniel@0:        case 'scoring_fn', scoring_fn = args{i+1};
Daniel@0:        case 'type',       type = args{i+1}; 
Daniel@0:        case 'discrete',   discrete = args{i+1}; 
Daniel@0:        case 'clamped',    clamped = args{i+1}; 
Daniel@0:        case 'params',     if isempty(args{i+1}), params = cell(1,n); else params = args{i+1};  end
Daniel@0:       end
Daniel@0:     end
Daniel@0:   else
Daniel@0:     max_fan_in = args{1};
Daniel@0:   end
Daniel@0: end
Daniel@0: 
Daniel@0: dag = zeros(n,n);
Daniel@0: 
Daniel@0: for i=1:n
Daniel@0:   ps = [];
Daniel@0:   j = order(i);
Daniel@0:   u = find(clamped(j,:)==0);    
Daniel@0:   score = score_family(j, ps, type{j}, scoring_fn, ns, discrete, data(:,u), params{j});
Daniel@0:   if verbose, fprintf('\nnode %d, empty score %6.4f\n', j, score); end
Daniel@0:   done = 0;
Daniel@0:   while ~done & (length(ps) <= max_fan_in)
Daniel@0:     pps = mysetdiff(order(1:i-1), ps); % potential parents
Daniel@0:     nps = length(pps);
Daniel@0:     pscore = zeros(1, nps);
Daniel@0:     for pi=1:nps
Daniel@0:       p = pps(pi);
Daniel@0:       pscore(pi) = score_family(j, [ps p], type{j}, scoring_fn, ns, discrete, data(:,u), params{j});
Daniel@0:       if verbose, fprintf('considering adding %d to %d, score %6.4f\n', p, j, pscore(pi)); end
Daniel@0:     end
Daniel@0:     [best_pscore, best_p] = max(pscore);
Daniel@0:     best_p = pps(best_p);
Daniel@0:     if best_pscore > score
Daniel@0:       score = best_pscore;
Daniel@0:       ps = [ps best_p];
Daniel@0:       if verbose, fprintf('* adding %d to %d, score %6.4f\n', best_p, j, best_pscore); end
Daniel@0:     else
Daniel@0:       done = 1;
Daniel@0:     end
Daniel@0:   end
Daniel@0:   if ~isempty(ps) % need this check for matlab 5.2
Daniel@0:     dag(ps, j) = 1;
Daniel@0:   end
Daniel@0: end
Daniel@0: 
Daniel@0: 
Daniel@0: 
Daniel@0: