diff options
Diffstat (limited to 'sourcecodes/bnt-master')
13 files changed, 247 insertions, 0 deletions
diff --git a/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_pot_orig.m b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_pot_orig.m new file mode 100644 index 00000000..dadcb033 --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_pot_orig.m @@ -0,0 +1,62 @@ +function pot = convert_to_pot_orig(CPD, pot_type, domain, evidence, n, ns_current) +% CONVERT_TO_POT Convert discrete CPD with original distribution to a potential +% pot = convert_to_pot_orig(CPD, pot_type, domain, evidence, n, ns_current) +% +% pots = CPD evaluated using evidence(domain) + +ncases = size(domain,2); +assert(ncases==1); % not yet vectorized + +sz = dom_sizes(CPD); +ns = zeros(1, max(domain)); +ns(domain) = sz; + +CPT1 = CPD_to_CPT(CPD); +spar = issparse(CPT1); +odom = domain(~isemptycell(evidence(domain))); +if spar + T = convert_to_sparse_table(CPD, domain, evidence); +else + T = convert_to_table(CPD, domain, evidence); +end + +switch pot_type + case 'u', + pot = upot(domain, sz, T, 0*myones(sz)); + case 'd', + CPT_orig = CPD_to_CPT_orig(CPD); + pot = dpot(n, ns_current, CPT_orig); + case {'c','g'}, + % Since we want the output to be a Gaussian, the whole family must be observed. + % In other words, the potential is really just a constant. + p = T; + %p = prob_node(CPD, evidence(domain(end)), evidence(domain(1:end-1))); + ns(domain) = 0; + pot = cpot(domain, ns(domain), log(p)); + + case 'cg', + T = T(:); + ns(odom) = 1; + can = cell(1, length(T)); + for i=1:length(T) + if T(i) == 0 + can{i} = cpot([], [], -Inf); % bug fix by Bob Welch 20/2/04 + else + can{i} = cpot([], [], log(T(i))); + end; + end + pot = cgpot(domain, [], ns, can); + + case 'scg' + T = T(:); + ns(odom) = 1; + pot_array = cell(1, length(T)); + for i=1:length(T) + pot_array{i} = scgcpot([], [], T(i)); + end + pot = scgpot(domain, [], [], ns, pot_array); + + otherwise, + error(['unrecognized pot type ' pot_type]) +end + diff --git a/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_pot_orig.m~ b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_pot_orig.m~ new file mode 100644 index 00000000..ecc57d49 --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_pot_orig.m~ @@ -0,0 +1,62 @@ +function pot = convert_to_pot(CPD, pot_type, domain, evidence) +% CONVERT_TO_POT Convert a discrete CPD to a potential +% pot = convert_to_pot(CPD, pot_type, domain, evidence) +% +% pots = CPD evaluated using evidence(domain) + +ncases = size(domain,2); +assert(ncases==1); % not yet vectorized + +sz = dom_sizes(CPD); +ns = zeros(1, max(domain)); +ns(domain) = sz; + +CPT1 = CPD_to_CPT(CPD); +spar = issparse(CPT1); +odom = domain(~isemptycell(evidence(domain))); +if spar + T = convert_to_sparse_table(CPD, domain, evidence); +else + T = convert_to_table(CPD, domain, evidence); +end + +switch pot_type + case 'u', + pot = upot(domain, sz, T, 0*myones(sz)); + case 'd', + ns(odom) = 1; + pot = dpot(domain, ns(domain), T); + case {'c','g'}, + % Since we want the output to be a Gaussian, the whole family must be observed. + % In other words, the potential is really just a constant. + p = T; + %p = prob_node(CPD, evidence(domain(end)), evidence(domain(1:end-1))); + ns(domain) = 0; + pot = cpot(domain, ns(domain), log(p)); + + case 'cg', + T = T(:); + ns(odom) = 1; + can = cell(1, length(T)); + for i=1:length(T) + if T(i) == 0 + can{i} = cpot([], [], -Inf); % bug fix by Bob Welch 20/2/04 + else + can{i} = cpot([], [], log(T(i))); + end; + end + pot = cgpot(domain, [], ns, can); + + case 'scg' + T = T(:); + ns(odom) = 1; + pot_array = cell(1, length(T)); + for i=1:length(T) + pot_array{i} = scgcpot([], [], T(i)); + end + pot = scgpot(domain, [], [], ns, pot_array); + + otherwise, + error(['unrecognized pot type ' pot_type]) +end + diff --git a/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_table_orig.m b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_table_orig.m new file mode 100644 index 00000000..92aacc0d --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_table_orig.m @@ -0,0 +1,15 @@ +function T = convert_to_table(CPD, domain, evidence) +% CONVERT_TO_TABLE Convert a discrete CPD to a table with original distribution +% T = convert_to_table(CPD, domain, evidence) +% +% We convert the CPD to a CPT, and then lookup the evidence on the discrete parents. +% The resulting table can easily be converted to a potential. + +domain = domain(:); +CPT = CPD_to_CPT_orig(CPD); +odom = domain(~isemptycell(evidence(domain))); +vals = cat(1, evidence{odom}); +map = find_equiv_posns(odom, domain); +index = mk_multi_index(length(domain), map, vals); +T = CPT(index{:}); +T = T(:); diff --git a/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_table_orig.m~ b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_table_orig.m~ new file mode 100644 index 00000000..dc5bcd40 --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@discrete_CPD/convert_to_table_orig.m~ @@ -0,0 +1,15 @@ +function T = convert_to_table(CPD, domain, evidence) +% CONVERT_TO_TABLE Convert a discrete CPD to a table +% T = convert_to_table(CPD, domain, evidence) +% +% We convert the CPD to a CPT, and then lookup the evidence on the discrete parents. +% The resulting table can easily be converted to a potential. + +domain = domain(:); +CPT = CPD_to_CPT(CPD); +odom = domain(~isemptycell(evidence(domain))); +vals = cat(1, evidence{odom}); +map = find_equiv_posns(odom, domain); +index = mk_multi_index(length(domain), map, vals); +T = CPT(index{:}); +T = T(:); diff --git a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/CPD_to_CPT_orig.m b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/CPD_to_CPT_orig.m new file mode 100644 index 00000000..707de900 --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/CPD_to_CPT_orig.m @@ -0,0 +1,5 @@ +function CPT = CPD_to_CPT_orig(CPD) +% CPD_TO_CPT Convert the discrete CPD to tabular form (tabular) +% CPT = CPD_to_CPT_orig(CPD) + +CPT = CPD.CPT_orig; diff --git a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/CPD_to_CPT_orig.m~ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/CPD_to_CPT_orig.m~ new file mode 100644 index 00000000..351f103c --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/CPD_to_CPT_orig.m~ @@ -0,0 +1,5 @@ +function CPT = CPD_to_CPT(CPD) +% CPD_TO_CPT Convert the discrete CPD to tabular form (tabular) +% CPT = CPD_to_CPT(CPD) + +CPT = CPD.CPT; diff --git a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/get_field.m b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/get_field.m index ba233db9..5812330d 100644 --- a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/get_field.m +++ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/get_field.m @@ -10,6 +10,7 @@ function val = get_field(CPD, name) switch name case 'cpt', val = CPD.CPT; + case 'cpt_orig', val = CPD.CPT_orig; case 'counts', val = CPD.counts; otherwise, error(['invalid argument name ' name]); diff --git a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/learn_params_orig.m b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/learn_params_orig.m new file mode 100644 index 00000000..c0948566 --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/learn_params_orig.m @@ -0,0 +1,20 @@ +function CPD = learn_params_orig(CPD,j,data,ns,cnodes) +% LEARN_PARAMS_ORIG +% Calculate the original distributions of the data. +% The original distributions are just the percentages of states in the +% data file. + +local_data = data(j, :); +nobs = size(local_data,2); +if iscell(local_data) + local_data = cell2num(local_data); +end +counts = compute_counts(local_data,ns(j)); +counts = counts/nobs; +switch CPD.prior_type + case 'none', CPD.CPT_orig = counts; +% case 'dirichlet', CPD.CPT = mk_stochastic(counts + CPD.dirichlet); +% I will use 'dirichlet' priors incorrectly here. + case 'dirichlet', CPD.CPT_orig = counts; + otherwise, error(['unrecognized prior ' CPD.prior_type]) +end diff --git a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/learn_params_orig.m~ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/learn_params_orig.m~ new file mode 100644 index 00000000..7a19a42d --- /dev/null +++ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/learn_params_orig.m~ @@ -0,0 +1,20 @@ +function CPD = learn_params_orig(CPD,j,data,data,ns,cnodes) +% LEARN_PARAMS_ORIG +% Calculate the original distributions of the data. +% The original distributions are just the percentages of states in the +% data file. + +local_data = data(j, :); +nobs = size(local_data,2); +if iscell(local_data) + local_data = cell2num(local_data); +end +counts = compute_counts(local_data,ns(j)); +counts = counts/nobs; +switch CPD.prior_type + case 'none', CPD.CPT_orig = counts; +% case 'dirichlet', CPD.CPT = mk_stochastic(counts + CPD.dirichlet); +% I will use 'dirichlet' priors incorrectly here. + case 'dirichlet', CPD.CPT_orig = counts; + otherwise, error(['unrecognized prior ' CPD.prior_type]) +end diff --git a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/tabular_CPD.m b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/tabular_CPD.m index 728302d4..a41a23d9 100644 --- a/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/tabular_CPD.m +++ b/sourcecodes/bnt-master/BNT/CPDs/@tabular_CPD/tabular_CPD.m @@ -9,6 +9,7 @@ function CPD = tabular_CPD(bnet, self, varargin) % - T means use table T; it will be reshaped to the size of node's family. % - 'rnd' creates rnd params (drawn from uniform) % - 'unif' creates a uniform distribution +% CPT_orig - specifies the distribution based on original data % adjustable - 0 means don't adjust the parameters during learning [1] % prior_type - defines type of prior ['none'] % - 'none' means do ML estimation @@ -60,6 +61,7 @@ CPD.sparse = 0; % set defaults CPD.CPT = mk_stochastic(myrand(fam_sz)); +CPD.CPT_orig = mk_stochastic(myrand(ns([self]))); CPD.adjustable = 1; CPD.prior_type = 'none'; dirichlet_type = 'BDeu'; @@ -158,6 +160,7 @@ function CPD = init_fields() % or create it from scratch. (Matlab requires this.) CPD.CPT = []; +CPD.CPT_orig = []; CPD.sizes = []; CPD.prior_type = []; CPD.dirichlet = []; diff --git a/sourcecodes/bnt-master/BNT/inference/static/@jtree_inf_engine/marginal_nodes_no_ev.m b/sourcecodes/bnt-master/BNT/inference/static/@jtree_inf_engine/marginal_nodes_no_ev.m new file mode 100644 index 00000000..220690ce --- /dev/null +++ b/sourcecodes/bnt-master/BNT/inference/static/@jtree_inf_engine/marginal_nodes_no_ev.m @@ -0,0 +1,18 @@ +function marginal = marginal_nodes_no_ev(bnet,engine, query) +% MARGINAL_NODES Get original distribution of the specified query nodes (jtree) +% marginal = marginal_nodes(bnet, engine, query) +% + +if ismember(query,bnet.dnodes) + marginal.domain = query; + marginal.T = CPD_to_CPT_orig(bnet.CPD{query}); + marginal.mu = []; + marginal.Sigma = []; +else + c = clq_containing_nodes(engine, query); + if c == -1 + error(['no clique contains ' num2str(query)]); + end + marginal = pot_to_marginal(marginalize_pot(engine.clpot{c}, query, engine.maximize)); +end + diff --git a/sourcecodes/bnt-master/BNT/inference/static/@jtree_inf_engine/marginal_nodes_no_ev.m~ b/sourcecodes/bnt-master/BNT/inference/static/@jtree_inf_engine/marginal_nodes_no_ev.m~ new file mode 100644 index 00000000..220690ce --- /dev/null +++ b/sourcecodes/bnt-master/BNT/inference/static/@jtree_inf_engine/marginal_nodes_no_ev.m~ @@ -0,0 +1,18 @@ +function marginal = marginal_nodes_no_ev(bnet,engine, query) +% MARGINAL_NODES Get original distribution of the specified query nodes (jtree) +% marginal = marginal_nodes(bnet, engine, query) +% + +if ismember(query,bnet.dnodes) + marginal.domain = query; + marginal.T = CPD_to_CPT_orig(bnet.CPD{query}); + marginal.mu = []; + marginal.Sigma = []; +else + c = clq_containing_nodes(engine, query); + if c == -1 + error(['no clique contains ' num2str(query)]); + end + marginal = pot_to_marginal(marginalize_pot(engine.clpot{c}, query, engine.maximize)); +end + diff --git a/sourcecodes/bnt-master/BNT/learning/learn_params.m b/sourcecodes/bnt-master/BNT/learning/learn_params.m index 1bf9c843..3d7ba847 100644 --- a/sourcecodes/bnt-master/BNT/learning/learn_params.m +++ b/sourcecodes/bnt-master/BNT/learning/learn_params.m @@ -21,6 +21,9 @@ for j=1:n fam = family(bnet.dag,j); %bnet.CPD{j} = learn_params(bnet.CPD{j}, data(fam,:)); bnet.CPD{j} = learn_params(bnet.CPD{j}, fam, data, bnet.node_sizes, bnet.cnodes); + if ismember(e,bnet.dnodes) + bnet.CPD{j} = learn_params_orig(bnet.CPD{j}, j, data, bnet.node_sizes, bnet.cnodes); + end end end |