Information about the BattMo tests

Octave vs MATLAB

The tests only run in MATLAB.

Running the tests (simple commands)

The tests are run using the command `runtests`. To run all the tests in path, run runtests.

To run all the tests in the folder ‘Tests’, run runtests('Tests', includeSubfolders, true).

To only run the test TestBattery1D, run runtests('TestBattery1D').

Running the tests (some more advanced commands)

To stop on first failure, run

import matlab.unittest.TestSuite;
import matlab.unittest.TestRunner
import matlab.unittest.plugins.StopOnFailuresPlugin
suite = testsuite('TestBattery1D');
runner = testrunner('textoutput');
runner.addPlugin(StopOnFailuresPlugin)
results = runner.run(suite);

The results of the tests come as TestResult classes

To run in parallel

import matlab.unittest.TestSuite;
import matlab.unittest.TestRunner
suite = TestSuite.fromClass(?TestBattery1D);
runner = TestRunner.withNoPlugins();
result = runner.runInParallel(suite);
result.table()

In the last line the results are converted to a table for easier visualization.

Note that, when running in parallel, we need to setup the mrst modules before (call to mrstModule add modulname). It appears that calling mrstModule inside TestCase constructor does not work.

We can filter the suite by given the value of a parameter. For example, in the following, we run only the short versions for the tests (only a few timesteps).

import matlab.unittest.TestSuite;
import matlab.unittest.selectors.HasParameter;
suite = TestSuite.fromClass(?TestBattery1D);
suite = suite.selectIf(HasParameter('Property', 'testSize', 'Value', 'short'));
suite.run();

We can also use matlab standard way (no selector), as in the following example

import matlab.unittest.TestSuite;
suite = TestSuite.fromClass(?TestRunExamples);
names = arrayfun(@(test) test.Parameterization.Value, suite, 'un', false);
ind = ismember(names, {'runBattery1D', ...
                       'runBattery2D'    , ...
                       'runBattery3D'    , ...
                       'runBattery2DMech', ...
                       'runChen2020'});
suite = suite(ind);
suite.run()

Through plugins, we can add extra output in the tests

import matlab.unittest.TestSuite;
import matlab.unittest.TestRunner
import matlab.unittest.plugins.*
suite = TestSuite.fromClass(?TestBattery1D);
runner = TestRunner.withNoPlugins();
outputstream = ToFile('report.txt');
runner.addPlugin(DiagnosticsOutputPlugin(outputstream));
runner.addPlugin(DiagnosticsRecordingPlugin('IncludingPassingDiagnostics',true));
result = runner.run(suite);

Then, look at structure result(1).Details.DiagnosticRecord and the file report.txt

Generate Reference Data

Reference data is created by setting the property createReferenceData to true as an external parameter,

import matlab.unittest.TestSuite;
import matlab.unittest.parameters.Parameter;
param = Parameter.fromData("createReferenceData", {true});
suite = TestSuite.fromFolder('TestExamples', 'ExternalParameters', param);
result = suite.run()

Reference data are then saved in mat files found in TestExamples/ReferenceData.