wolffd@0: function write_mat_results_ismir12(filein, subrun, fileout)
wolffd@0: % write_mat_resutls_ismir12(filein, subrun, fileout)
wolffd@0: %
wolffd@0: % write results from result file into ismir12 generic format.
wolffd@0: % if subrun is a vector, the results are averaged over the 
wolffd@0: % runs specified
wolffd@0: 
wolffd@0: 
wolffd@0: [out, stats, features, individual] = test_generic_display_results(filein, 0);
wolffd@0: 
wolffd@0: % ---
wolffd@0: % we get the single individual values and 
wolffd@0: % ---
wolffd@0: for i = 1:numel(subrun)
wolffd@0:     individual = individual(subrun);
wolffd@0:     
wolffd@0:     % number of inctrain cycles
wolffd@0:     n_inctrain = numel(individual.diag.inctrain);
wolffd@0: 
wolffd@0:     % ---
wolffd@0:     % ok_train_unused
wolffd@0:     % ---
wolffd@0:     values_ok_train_unused{i} = reshape([individual.diag.inctrain.ok_notin_train], [],n_inctrain);
wolffd@0:     values_ok_train_unused{i} = values_ok_train_unused(1:2:end,:).*100;
wolffd@0: 
wolffd@0:     mean_ok_train_unused{i} = mean(values_ok_train_unused, 1);
wolffd@0:     var_ok_train_unused{i} = var(values_ok_train_unused,[], 1);
wolffd@0: 
wolffd@0:     % ---
wolffd@0:     % ok_train
wolffd@0:     % ---
wolffd@0:     values_ok_train{i} = reshape([individual.diag.inctrain.ok_train], [],n_inctrain);
wolffd@0:     values_ok_train{i} = values_ok_train(1:2:end,:).*100;
wolffd@0: 
wolffd@0:     mean_ok_train{i} = mean(values_ok_train, 1);
wolffd@0:     var_ok_train{i} = var(values_ok_train,[], 1);
wolffd@0: 
wolffd@0:     % ---
wolffd@0:     % ok_test
wolffd@0:     % ---
wolffd@0:     values_ok_test{i} = reshape([individual.diag.inctrain.ok_test], [],n_inctrain);
wolffd@0:     values_ok_test{i} = values_ok_test(1:2:end,:).*100;
wolffd@0: 
wolffd@0:     mean_ok_test{i} = mean(values_ok_test, 1);
wolffd@0:     var_ok_test{i} = var(values_ok_test,[], 1);
wolffd@0: end
wolffd@0: 
wolffd@0: 
wolffd@0: 
wolffd@0: clear ('out', 'stats', 'features', 'individual');
wolffd@0: save(fileout)