| name | matlab-test-generator |
| description | Create comprehensive MATLAB unit tests using the MATLAB Testing Framework. Use when generating test files, test cases, unit tests, test suites, or when the user requests testing for MATLAB code, functions, or classes. |
| license | MathWorks BSD-3-Clause (see LICENSE) |
| allowed-tools | Read, Write, Edit, Bash, Grep, Glob |
MATLAB Test Generator
This skill provides comprehensive guidelines for creating robust unit tests using the MATLAB Testing Framework. Generate test classes, test methods, and test suites following MATLAB best practices.
When to Use This Skill
- Creating unit tests for MATLAB functions or classes
- Generating test suites for existing code
- Writing test cases with assertions and fixtures
- Creating parameterized tests
- Setting up test environments with setup/teardown methods
- When user requests testing, test coverage, or mentions unit tests
MATLAB Testing Framework Overview
MATLAB supports multiple testing approaches:
- Script-Based Tests - Simple test scripts with assertions
- Function-Based Tests - Test functions with local functions for each test
- Class-Based Tests - Full-featured test classes (recommended for complex testing)
Test Class Structure
Basic Test Class Template
classdef MyFunctionTest < matlab.unittest.TestCase
% Tests for myFunction
properties (TestParameter)
% Define parameters for parameterized tests
end
properties
% Test fixtures and shared data
end
methods (TestClassSetup)
% Runs once before all tests
end
methods (TestClassTeardown)
% Runs once after all tests
end
methods (TestMethodSetup)
% Runs before each test method
end
methods (TestMethodTeardown)
% Runs after each test method
end
methods (Test)
% Individual test methods
end
end
Critical Rules
File Naming
- Test files MUST end with
Test.m(e.g.,myFunctionTest.m) - Class name must match filename
- Follow PascalCase naming convention
- Place in
tests/directory or alongside source withTestsuffix
Test Method Naming
- Use descriptive, readable names
- Start with lowercase letter
- Use camelCase
- Describe what is being tested
- Example:
testAdditionWithPositiveNumbers
Assertions
Always use the appropriate assertion method:
verifyEqual(actual, expected)- Continues testing on failureassertEqual(actual, expected)- Stops testing on failure (legacy)verifyTrue(condition)- Verify boolean conditionverifyError(f, errorID)- Verify function throws errorverifyWarning(f, warningID)- Verify function issues warningverifyInstanceOf(actual, expectedClass)- Verify object typeverifySize(actual, expectedSize)- Verify array sizeverifyEmpty(actual)- Verify empty array
Complete Test Examples
Example 1: Testing a Simple Function
Function to test (add.m):
function result = add(a, b)
% ADD Add two numbers
result = a + b;
end
Test class (addTest.m):
classdef addTest < matlab.unittest.TestCase
% Tests for add function
methods (Test)
function testAddPositiveNumbers(testCase)
% Test addition of positive numbers
result = add(2, 3);
testCase.verifyEqual(result, 5);
end
function testAddNegativeNumbers(testCase)
% Test addition of negative numbers
result = add(-2, -3);
testCase.verifyEqual(result, -5);
end
function testAddMixedNumbers(testCase)
% Test addition of positive and negative
result = add(5, -3);
testCase.verifyEqual(result, 2);
end
function testAddZero(testCase)
% Test addition with zero
result = add(5, 0);
testCase.verifyEqual(result, 5);
end
function testAddArrays(testCase)
% Test element-wise addition of arrays
result = add([1 2 3], [4 5 6]);
testCase.verifyEqual(result, [5 7 9]);
end
end
end
Example 2: Parameterized Tests
classdef calculateAreaTest < matlab.unittest.TestCase
% Tests for calculateArea function with parameters
properties (TestParameter)
rectangleDims = struct(...
'small', struct('width', 2, 'height', 3, 'expected', 6), ...
'large', struct('width', 10, 'height', 20, 'expected', 200), ...
'unit', struct('width', 1, 'height', 1, 'expected', 1));
end
methods (Test)
function testRectangleArea(testCase, rectangleDims)
% Test rectangle area calculation
area = calculateArea(rectangleDims.width, rectangleDims.height);
testCase.verifyEqual(area, rectangleDims.expected);
end
end
end
Example 3: Tests with Fixtures
classdef DataProcessorTest < matlab.unittest.TestCase
% Tests for DataProcessor class with setup/teardown
properties
TestData
TempFile
end
methods (TestMethodSetup)
function createTestData(testCase)
% Create test data before each test
testCase.TestData = rand(100, 5);
testCase.TempFile = tempname;
end
end
methods (TestMethodTeardown)
function deleteTestData(testCase)
% Clean up after each test
if isfile(testCase.TempFile)
delete(testCase.TempFile);
end
end
end
methods (Test)
function testDataLoading(testCase)
% Test data loading functionality
save(testCase.TempFile, 'data', testCase.TestData);
processor = DataProcessor(testCase.TempFile);
testCase.verifyEqual(processor.data, testCase.TestData);
end
function testDataNormalization(testCase)
% Test data normalization
processor = DataProcessor();
processor.data = testCase.TestData;
normalized = processor.normalize();
testCase.verifyLessThanOrEqual(max(normalized(:)), 1);
testCase.verifyGreaterThanOrEqual(min(normalized(:)), 0);
end
end
end
Example 4: Error Testing
classdef validateInputTest < matlab.unittest.TestCase
% Tests for validateInput function
methods (Test)
function testValidInput(testCase)
% Test valid input passes
testCase.verifyWarningFree(@() validateInput(5));
end
function testNegativeInputError(testCase)
% Test negative input throws error
testCase.verifyError(@() validateInput(-5), 'MATLAB:validators:mustBePositive');
end
function testNonNumericInputError(testCase)
% Test non-numeric input throws error
testCase.verifyError(@() validateInput('text'), 'MATLAB:validators:mustBeNumeric');
end
function testEmptyInputError(testCase)
% Test empty input throws error
testCase.verifyError(@() validateInput([]), 'MATLAB:validators:mustBeNonempty');
end
end
end
Test Organization Best Practices
Directory Structure
project/
├── src/
│ ├── myFunction.m
│ └── MyClass.m
└── tests/
├── myFunctionTest.m
└── MyClassTest.m
Running Tests
Run all tests in directory:
results = runtests('tests')
Run specific test class:
results = runtests('myFunctionTest')
Run specific test method:
results = runtests('myFunctionTest/testAddPositiveNumbers')
Generate test suite:
suite = testsuite('tests');
results = run(suite);
With coverage report:
import matlab.unittest.TestRunner
import matlab.unittest.plugins.CodeCoveragePlugin
runner = TestRunner.withTextOutput;
runner.addPlugin(CodeCoveragePlugin.forFolder('src'));
results = runner.run(testsuite('tests'));
Assertion Tolerance
For floating-point comparisons, use tolerance:
% Absolute tolerance
testCase.verifyEqual(actual, expected, 'AbsTol', 1e-10);
% Relative tolerance
testCase.verifyEqual(actual, expected, 'RelTol', 1e-6);
% Both
testCase.verifyEqual(actual, expected, 'AbsTol', 1e-10, 'RelTol', 1e-6);
Test Assumptions
Use assumptions to skip tests when prerequisites aren't met:
function testPlotting(testCase)
% Skip test if not running in graphical environment
testCase.assumeTrue(usejava('desktop'), 'Requires desktop environment');
% Test code here
fig = figure;
plot(1:10);
testCase.verifyInstanceOf(fig, 'matlab.ui.Figure');
close(fig);
end
Test Tagging
Tag tests for selective execution:
methods (Test, TestTags = {'Unit'})
function testBasicOperation(testCase)
% Fast unit test
end
end
methods (Test, TestTags = {'Integration', 'Slow'})
function testDatabaseConnection(testCase)
% Slower integration test
end
end
Run tagged tests:
% Run only Unit tests
suite = testsuite('tests', 'Tag', 'Unit');
results = run(suite);
Mock Objects and Stubs
For testing code with dependencies:
function testWithMock(testCase)
% Create mock object
import matlab.mock.TestCase
[mock, behavior] = testCase.createMock(?MyInterface);
% Define behavior
testCase.assignOutputsWhen(withAnyInputs(behavior.methodName), expectedOutput);
% Use mock in test
result = functionUnderTest(mock);
testCase.verifyEqual(result, expectedValue);
end
Performance Testing
Test execution time:
function testPerformance(testCase)
% Test function executes within time limit
tic;
result = expensiveFunction(largeData);
elapsedTime = toc;
testCase.verifyLessThan(elapsedTime, 1.0, ...
'Function should complete within 1 second');
end
Checklist for Test Creation
Before finalizing tests, verify:
- Test file ends with
Test.m - Class inherits from
matlab.unittest.TestCase - Test methods use descriptive names
- Appropriate assertions used (verify vs assert)
- Edge cases covered (empty, zero, negative, large values)
- Error cases tested with
verifyError - Floating-point comparisons use tolerance
- Setup/teardown methods clean up resources
- Tests are independent (can run in any order)
- Tests follow Arrange-Act-Assert pattern
- Parameterized tests used for similar test cases
- Comments explain what is being tested
Common Patterns
Arrange-Act-Assert Pattern
function testCalculation(testCase)
% Arrange - Set up test data
input1 = 5;
input2 = 3;
expected = 8;
% Act - Execute the code under test
actual = myFunction(input1, input2);
% Assert - Verify the result
testCase.verifyEqual(actual, expected);
end
Testing Private Methods
% Use access to test private methods
classdef MyClassTest < matlab.unittest.TestCase
methods (Test)
function testPrivateMethod(testCase)
obj = MyClass();
% Access private method
result = obj.privateMethod(testCase);
testCase.verifyEqual(result, expectedValue);
end
end
end
Testing Static Methods
function testStaticMethod(testCase)
% Test static method without instantiation
result = MyClass.staticMethod(input);
testCase.verifyEqual(result, expected);
end
Troubleshooting
Issue: Tests not discovered
- Solution: Ensure filename ends with
Test.mand class inherits frommatlab.unittest.TestCase
Issue: Floating-point comparison failures
- Solution: Use
'AbsTol'or'RelTol'parameters withverifyEqual
Issue: Tests affecting each other
- Solution: Ensure proper cleanup in teardown methods and test independence
Issue: Slow test execution
- Solution: Use tags to separate fast unit tests from slow integration tests
Additional Resources
- Use
doc matlab.unittest.TestCasefor complete assertion reference - Use
doc matlab.unittest.fixturesfor advanced fixture usage - Use
doc matlab.mockfor mocking framework documentation