format output of tests

src/BeatTracker.m

@@ -104,7 +104,7 @@ function init_model(obj)
         function init_train_data(obj)
             % set up the training data and load or compute the cluster
             % assignments of the bars/beats
-            fprintf('* Set up training data ...\n');
+            fprintf('* Set up training data\n');
             % check if features are already saved
             if isfield(obj.Params, 'stored_train_data_fln') && ...
                     exist(obj.Params.stored_train_data_fln, 'file') && ...
@@ -156,7 +156,7 @@ function init_train_data(obj)
                 obj.train_data = obj.train_data.read_pattern_bars(...
                     obj.Params.clusterIdFln, obj.Params.pattern_size);
             else
-                fprintf('* Clustering data by %s ...\n', ...
+                fprintf('* Clustering data by %s: ', ...
                     obj.Params.cluster_type);
                 if ismember(obj.Params.cluster_type, {'meter', ...
                         'rhythm'})
@@ -180,7 +180,7 @@ function init_train_data(obj)
                             'plotting_path', obj.Params.results_path);
                     end
                 end
-                fprintf('done\n  %i clusters detected.\n', ...
+                fprintf('  %i clusters detected.\n', ...
                     obj.train_data.clustering.n_clusters);
             end
             obj.Params.R = obj.train_data.clustering.n_clusters;
@@ -198,7 +198,7 @@ function init_test_data(obj)
         end
 
         function train_model(obj)
-            fprintf('* Training model (%s) ...\n',...
+            fprintf('* Training model (%s)\n',...
                 obj.Params.inferenceMethod)
             if isempty(obj.init_model_fln)
                 obj.model.train_model(obj.Params.transition_params, ...
@@ -233,7 +233,7 @@ function train_model(obj)
 
         function results = do_inference(obj, test_file_id)
             [~, fname, ~] = fileparts(obj.test_data.file_list{test_file_id});
-            fprintf('* Started inference on %s', fname);
+            fprintf('* Inferring meter from %s\n', fname);
             % load feature
             observations = obj.feature.load_feature(...
                 obj.test_data.file_list{test_file_id}, ...

src/BeatTrackingModelHMM.m

@@ -132,7 +132,7 @@ function make_initial_distribution(obj, train_data)
                     do_output, fname, y);
             elseif strcmp(inference_method, 'HMM_viterbi')
                 % decode MAP state sequence using Viterbi
-                fprintf('* Decoding (viterbi) .');
+                fprintf('* Decoding (viterbi)');
                 if exist('belief_func', 'var')
                     % use viterbi with belief functions
                     hidden_state_sequence = obj.viterbi_iteration(obs_lik, ...

src/ParticleFilter.m

@@ -25,7 +25,6 @@
             debug = 0;
             % initialize particles and preallocate memory
             n_frames = size(obs_lik, 3);
-            perc = round(0.1*n_frames);
             % add one frame, which corresponds to the initial distribution
             % at time 0 before the first observation comes in
             m = zeros(obj.n_particles, n_frames + 1, 'single');
@@ -66,11 +65,7 @@
                     logP_data_pf(:, 4, iFrame) = w;
                     logP_data_pf(:, 5, iFrame) = g;
                 end
-                if rem(iFrame, perc) == 0
-                    fprintf('.');
-                end
             end
-            fprintf('\n');
             % remove initial state
             m = m(:, 2:end); n = n(:, 2:end); r = r(:, 2:end);
             if debug, save('/tmp/data_pf.mat', 'logP_data_pf', 'obs_lik'); end

tests/ExamplesTest.m

@@ -11,6 +11,7 @@
     methods (Test)
         function testEx1(testCase)
             addpath(testCase.example_path)
+            fprintf('\ntestEx1: \n');
             % Beat tracking with a pre-trained HMM
             Results = ex1_beat_tracking_with_pretrained_hmm(...
                 testCase.audio_files{1}, testCase.result_folder);
@@ -22,6 +23,7 @@ function testEx1(testCase)
 
         function testEx2(testCase)
             addpath(testCase.example_path)
+            fprintf('\ntestEx2: \n');
             % Beat tracking with a pre-trained PF
             % Initialize the random number generator using a seed of 1 
             % to make the results in this example repeatable
@@ -36,6 +38,7 @@ function testEx2(testCase)
 
         function testEx3(testCase)
             addpath(testCase.example_path)
+            fprintf('\ntestEx3: \n');
             rng('default'); rng(1);
             % Train a HMM and test it
             Results = ex3_train_and_test_hmm(testCase.audio_files{1}, ...
@@ -48,6 +51,7 @@ function testEx3(testCase)
 
         function testEx4(testCase)
             addpath(testCase.example_path)
+            fprintf('\ntestEx4: \n');
             rng('default'); rng(1);
             % Train a HMM and test it
             Results = ex4_train_and_test_hmm(testCase.audio_files{1}, ...
@@ -60,6 +64,7 @@ function testEx4(testCase)
 
         function testEx5(testCase)
             addpath(testCase.example_path)
+            fprintf('\ntestEx5: \n');
             rng('default'); rng(1);
             % Train a HMM and test it
             Results = ex5_train_and_test_pf(testCase.audio_files{1}, ...

tests/run_all_tests.m

@@ -5,6 +5,6 @@
 suiteFolder = TestSuite.fromFolder(pwd);
 % Add the bayesbeat class to the search path
 [func_path, ~, ~] = fileparts(mfilename('fullpath'));
-addpath(fullfile(func_path, '..'))
+addpath(fullfile(func_path, '../src'))
 % Test it
 result = run(suiteFolder);