matlab/matlab-agentic-toolkit / skills-catalog/matlab-software-development/matlab-write-tests
matlab-write-tests
VerifiedGenerate and structure MATLAB unit tests using matlab.unittest and matlab.uitest features, including class-based tests, parameterized testing, fixtures, mocking, and app testing with gestures. Use when writing, generating, or adding tests, creating test classes, adding test methods, parameterizing tests, setting up fixtures, mocking dependencies, or testing App Designer apps. Do NOT use for running tests, collecting coverage, or CI/CD configuration.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 11.2 KB |
| manifest.yaml | config | 495 B |
| references/app-testing-guidance.md | markdown | 7.3 KB |
| references/baseline-tests-guidance.md | markdown | 6.0 KB |
| references/constraints-guidance.md | markdown | 2.2 KB |
| references/fixtures-guidance.md | markdown | 2.7 KB |
| references/mocking-guidance.md | markdown | 3.3 KB |
| references/parameterized-tests-guidance.md | markdown | 3.6 KB |
references/parameterized-tests-guidance.md
View on GitHub# Parameterized Tests
Data-driven testing using `TestParameter` properties.
## Named Parameters with Structs
Use structs for clearer test names and grouped input/expected data:
```matlab
properties (TestParameter)
rectangleCase = struct(...
'small', struct('input', [2 3], 'expected', 6), ...
'large', struct('input', [10 20], 'expected', 200));
end
methods (Test)
function testRectangleArea(testCase, rectangleCase)
area = calculateArea(rectangleCase.input(1), rectangleCase.input(2));
testCase.verifyEqual(area, rectangleCase.expected);
end
end
```
## Parameter Combinations
| Mode | What it does | When to use |
|------|-------------|-------------|
| `exhaustive` (default) | Every combination of all parameters | 1-2 parameters, or when every combination genuinely matters |
| `sequential` | Zips parameters positionally | Input-output pairs that must stay aligned (same length required) |
| `pairwise` | All 2-way interactions, fewer tests | 3+ parameters where exhaustive is too many |
Watch for combinatorial explosion: 4 params × 5 values = 625 tests with `exhaustive`.
```matlab
% exhaustive (default)
methods (Test)
function testConversion(testCase, dataType, arrayShape)
end
end
% sequential — paired (same length required)
methods (Test, ParameterCombination = 'sequential')
function testPairedInputs(testCase, input, expected)
testCase.verifyEqual(myFunction(input), expected);
end
end
% pairwise — all 2-way interactions
methods (Test, ParameterCombination = 'pairwise')
function testMultipleFactors(testCase, dataType, solver, tolerance)
end
end
```
## Edge Case Parameters
Include edge cases relevant to the function's input domain:
```matlab
properties (TestParameter)
numericEdge = struct('zero', 0, 'negative', -1, 'large', 1e15, 'nan', NaN);
arrayEdge = struct('empty', [], 'scalar', 5, 'row', [1 2 3], 'column', [1; 2; 3]);
end
```
## Dynamic Parameters with TestParameterDefinition
Use when parameters should refresh each run (e.g., files in a folder) or depend on a higher-level parameter.
```matlab
properties (TestParameter)
dataFile % No default — populated by method below
end
methods (Static, TestParameterDefinition)
function dataFile = getDataFiles()
listing = dir('testdata/*.mat');
dataFile = struct();
for i = 1:numel(listing)
[~, name] = fileparts(listing(i).name);
dataFile.(matlab.lang.makeValidName(name)) = ...
fullfile(listing(i).folder, listing(i).name);
end
end
end
```
### Parameters Dependent on ClassSetupParameter
```matlab
properties (ClassSetupParameter)
precision = {'single', 'double'};
end
properties (TestParameter)
tolerance % depends on precision
end
methods (Static, TestParameterDefinition)
function tolerance = getTolerance(precision)
toleranceMap.single = struct('loose', 1e-4, 'tight', 1e-6);
toleranceMap.double = struct('loose', 1e-10, 'tight', 1e-14);
tolerance = toleranceMap.(precision);
end
end
```
## Best Practices
1. **Parameterize only when assertion logic is identical** — if the test body needs `if`/`switch`, use separate methods
2. **Use meaningful names** — they appear in test results
3. **Group related values in structs** — keeps input/expected together
4. **Keep parameter tables small** — growing tables may mean the tests are testing different logic
5. **Use the right level** — `TestParameter` for test data, `ClassSetupParameter` for expensive shared fixtures
----
Copyright 2026 The MathWorks, Inc.
----
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |