MATLAB - Debugging

MATLAB - A Modern Approach to Debugging

Debugging is the systematic process of finding and fixing errors, or “bugs,” in your code. Modern MATLAB provides a powerful, integrated suite of tools to make this process efficient and intuitive. Mastering these tools is essential for writing robust and correct programs.

Errors typically fall into three categories:

• Syntax Errors: Mistakes in the language syntax, like a missing parenthesis or keyword. The MATLAB Editor catches most of these before you even run the code.
• Runtime Errors: Errors that occur while the code is running, such as trying to access a non-existent array element or dividing by zero. These cause your program to crash.
• Logical Errors: The most challenging type. The code runs without crashing but produces an incorrect result due to a flaw in your algorithm or logic.

Step 1: Proactive Debugging with the Editor and Code Analyzer

Your first line of defense is the MATLAB Editor itself. As you type, the integrated Code Analyzer constantly scans your code for problems.

• Syntax Errors: Marked with a red underline and a red square on the indicator bar on the right. Hovering over the underline provides a description of the error.
• Warnings & Improvements: Marked with an orange underline and an orange square. These are not fatal errors but highlight potential bugs, performance issues, or deviations from best practices (e.g., a variable that is created but never used).

Best Practice: Pay attention to the indicator bar. If it’s not green, investigate the warnings and errors before running your code. This proactive approach saves a lot of time.

Step 2: Interactive Debugging with Breakpoints

When logical errors occur, you need to inspect the state of your program as it runs. Breakpoints are markers that pause execution at a specific line, giving you full control.

Setting Breakpoints and Pausing Execution

To set a breakpoint, click in the gray gutter to the left of a line number in the Editor. A red circle will appear. When you run your code, execution will stop before that line is executed.

When paused:

• The command prompt changes from >> to K>>, indicating you are in debug mode.
• A green arrow points to the next line to be executed.
• You can inspect the values of variables by hovering your mouse over them in the Editor or by looking at the “Workspace” panel.

Controlling Execution Flow

Once paused, a new set of controls appears in the Editor tab ribbon:

• Continue: Resumes execution until the program ends or another breakpoint is hit.
• Step: Executes the current line and then pauses again on the very next line.
• Step In: If the current line is a function call, this moves the debugger into that function, allowing you to debug it line-by-line.
• Step Out: If you are inside a function, this executes the rest of the function and pauses at the line after the original function call.
• Stop: Terminates the program execution completely.

Step 3: Advanced and Programmatic Debugging

Conditional Breakpoints

If an error occurs inside a loop after many iterations, you don’t want to step through each one. Right-click a breakpoint to set a condition, for example i > 950. The code will only pause when that condition is true.

Automatic Breakpoints

You can instruct MATLAB to automatically enter debug mode whenever an error occurs. This is incredibly useful for finding the exact location of runtime errors.

% Type these commands in the Command Window

% Pause execution if any runtime error occurs
dbstop if error

% Pause if a warning is issued
dbstop if warning

% To turn them off later
dbclear if error

Step 4: Graceful Error Handling with try/catch

For production code, you don’t want your program to crash. You can anticipate and handle potential errors using a try/catch block.

try
    % Code that might cause an error
    data = readmatrix('non_existent_file.csv');
    disp('File read successfully!');
catch ME
    % This block runs only if an error occurs in the 'try' block
    warning('Could not read the file. Using default values instead.');
    fprintf('Error was: %s\n', ME.message); % Display the error message
    data = [1, 2, 3]; % Assign a default value
end
disp('Program continues...');

Step 5: Automated Code Analysis

MATLAB provides functions to programmatically check your code for issues, which is useful for automated testing and CI/CD pipelines.

Using the checkcode function

The checkcode function analyzes a file and displays a report of potential issues directly in the Command Window.

% Assume 'my_script.m' has a line like 'x = 5' without a semicolon
checkcode('my_script.m')

The output would be something like:

L 3 (C 5): Add a semicolon after the statement to hide the output (in a function).