Create Tests for MATLAB Source Code

Since R2024a

MATLAB^® Test™ enables you to interactively create tests for your scripts, functions, and classes in the Editor or Live Editor. The test file created for a source file contains a test class to exercise the specified source code. You can use the generated test code as a starting point for writing your own tests.

This example shows how to create a test for a source file and then build on the generated code.

Create Test File

With your source file open in the Editor or Live Editor, you can create a test file to test the script, function, or class in that source file. The created test file contains a test class to exercise the code in your source file. If your source file is a class definition file for a concrete class, then the test file contains code to exercise the public methods of that class, including its default or explicitly defined constructor. (since R2024b)

For example, in a file named quadraticSolver.m in your current folder, create the quadraticSolver function. The function takes as inputs the coefficients of a quadratic polynomial and returns the roots of that polynomial. To validate its inputs, the function uses an arguments block.

function r = quadraticSolver(a,b,c)
% quadraticSolver returns solutions to the
% quadratic equation a*x^2 + b*x + c = 0.

arguments
    a (1,1) {mustBeNumeric,mustBeNonzero}
    b (1,1) {mustBeNumeric}
    c (1,1) {mustBeNumeric}
end

r = sort([(-b + sqrt(b^2 - 4*a*c)) / (2*a); ...
    (-b - sqrt(b^2 - 4*a*c)) / (2*a)]);
end

To create a test file that contains initial test code for the quadraticSolver function, right-click in the Editor and select Create Test. (In MATLAB Online™, you can also create a test file from the Editor or Live Editor tab by clicking Generate Test in the Test section.)

This code provides the contents of the resulting test file. The generated test class:

• Includes autogenerated inputs that satisfy the arguments block in the quadraticSolver function

• Calls the quadraticSolver function to produce an output

• Uses the verifyEqual qualification method to test the produced output

% This is an autogenerated sample test for file quadraticSolver.m
classdef testquadraticSolver < matlab.unittest.TestCase

    methods (Test)

        function test_quadraticSolver(testCase)
            a = -7162.2732;
            b = -1564.7743;
            c = 8314.7105;

            % Specify the expected output(s) of
            % quadraticSolver
            expected_r = ;

            % Exercise the function quadraticSolver
            actual_r = quadraticSolver(a, b, c);

            testCase.verifyEqual(actual_r, expected_r);
        end
    end
end

Note

MATLAB Test attempts to generate input argument values for a function or class method only if that function or method uses an arguments block for input argument validation. If the function or method does not include an arguments block, then the test class does not include autogenerated data to represent the input arguments.

Build on Generated Test Code

A test file created for your script, function, or class serves only as a starting point. For example, you must at least specify the expected output in the test code generated for the quadraticSolver function before running the test. To build on the generated test code for your source file, use the code insertion options on the toolstrip to add methods and parameterization properties to your test class. You can choose whether to add a method or property at the test level, method-setup level, or class-setup level. For more information, see Insert Test Code Using Editor.

In this example, to finalize the test class generated for the quadraticSolver function, first complete the code of the generated Test method and then add another Test method using the code insertion options on the toolstrip.

Complete Generated Test Method

The generated test for the quadraticSolver function requires you to specify the expected output of the function. To complete the test and make it more readable, follow these steps:

1. Rename the test class as SolverTest and the Test method as solution. Then, remove the comments you no longer need from the file and save the file as SolverTest.m in your current folder.

2. Specify the expected output of the function by calling the roots function with the autogenerated values.

3. Because the verifyEqual method tests floating-point vectors, specify a tolerance for comparison.

classdef SolverTest < matlab.unittest.TestCase
    methods (Test)
        function solution(testCase)
            a = -7162.2732;
            b = -1564.7743;
            c = 8314.7105;
            expected_r = sort(roots([a b c]));
            actual_r = quadraticSolver(a,b,c);
            testCase.verifyEqual(actual_r,expected_r,AbsTol=eps)
        end
    end
end

Add Another Test Method

To test the quadraticSolver function against invalid inputs, add another Test method to the class by clicking in the Test section on the Editor tab.

Implement the added method by following these steps:

1. Rename the method as nonnumericInput.

2. Add code to the method to verify that the quadraticSolver function throws an error when it is called with inputs 1, '-3', and 2.

Save the file. This code provides the contents of the SolverTest class.

classdef SolverTest < matlab.unittest.TestCase
    methods (Test)
        function solution(testCase)
            a = -7162.2732;
            b = -1564.7743;
            c = 8314.7105;
            expected_r = sort(roots([a b c]));
            actual_r = quadraticSolver(a,b,c);
            testCase.verifyEqual(actual_r,expected_r,AbsTol=eps)
        end

        function nonnumericInput(testCase)
            testCase.verifyError(@()quadraticSolver(1,'-3',2), ...
                "MATLAB:validators:mustBeNumeric")
        end
    end
end

Run Tests in Test Class

You can run the tests in your test class interactively in the Editor or in the Test Browser app. For example, with the SolverTest code visible in the Editor, go to the Editor tab and in the Run section, click . In this example, both the tests pass.

For more information on how to run tests and customize your test run interactively, see Run Tests in Editor and Run Tests Using Test Browser.