runtests_single.m

% COPYRIGHT
% This file is part of ORCA: https://github.com/ayrna/orca
% Original authors: Pedro Antonio Gutiérrez, María Pérez Ortiz, Javier Sánchez Monedero
% Citation: If you use this code, please cite the associated paper http://www.uco.es/grupos/ayrna/orreview
% Copyright:
%     This software is released under the The GNU General Public License v3.0 licence
%     available at http://www.gnu.org/licenses/gpl-3.0.html
% RUN SINGLE TESTS
% This script runs single tests using the ORCA API. The toy dataset is used
% with a 1 -hold-out experimental. Reference performance is compared. To
% add an script just place it in the test/singletests folder following
% other examples
%
clear;
addpath Algorithms
addpath Measures
addpath Utils

tests_dir = ['tests' filesep 'singletests'];
eval(['addpath ' tests_dir])

files = dir([tests_dir filesep '*.m']);

% Load the different partitions of the dataset
load ../exampledata/1-holdout/toy/matlab/train_toy.0
load ../exampledata/1-holdout/toy/matlab/test_toy.0

% "patterns" refers to the input variables and targets to the output one
train.patterns = train_toy(:,1:end-1);
train.targets = train_toy(:,end);
test.patterns = test_toy(:,1:end-1);
test.targets = test_toy(:,end);

for i=1:length(files)
    % Clear (almost) all the variables to avoid issues with some mex files in Octave +4.2.2
    if (exist ('OCTAVE_VERSION', 'builtin') > 0)
      clear -x tests_dir files i train test
    else
		  clearvars -except tests_dir files i train test
    end
    fprintf('==============================================\n');
    cmd = files(i).name;
    fprintf('Processing %s...\n', cmd);
    eval(cmd(1:end-2))
end

% error is called in individual tests, so if code reach this point all the
% tests have been run successfully
fprintf('\nAll tests ended successfully\n')

rmpath Algorithms
rmpath Measures
rmpath Utils
eval(['rmpath ' tests_dir])