Convert for-Loops into parfor-Loops

R2026b

In some cases, you must modify the code to convert for-loops to parfor-loops. This example shows how to diagnose and fix parfor-loop problems using a simple nested for-loop. Run this code in MATLAB^® and examine the results.

for x = 0:0.1:1
    for y = 2:10
        A(y) = A(y-1) + y;
    end
end

To speed up the code, try to convert the for-loops to parfor-loops. This code produces errors.

parfor x = 0:0.1:1
    parfor y = 2:10
        A(y) = A(y-1) + y;
    end
end

In this case you cannot simply convert the for-loops to parfor-loops without modification. To make this work, you must change the code in several places. To diagnose the problems, copy this code into the MATLAB editor and look for Code Analyzer messages. Use the following steps to diagnose and resolve them.

Tip

If you encounter warnings or errors related to parfor-loop requirements, guidelines, or limitations in MATLAB Code Analyzer, you can use a MATLAB Copilot action to help diagnose the issue. In the Code Analyzer message, click the button to ask Copilot to explain the specific requirement or limitation that the code violates and suggest code changes that can make the loop valid for parallel execution. For an example workflow, see Resolve Issues in parfor-loops Using MATLAB Copilot.

1. Identify structural errors reported by Code Analyzer.

   When MATLAB analyzes the nested parfor-loops, Code Analyzer reports errors related to loop structure and loop variables:

   • The outer loop uses a noninteger step size, which is not supported by parfor.

   • A parfor-loop is nested inside another parfor-loop, which is not allowed.

   

   At this stage, Code Analyzer does not report any variable classification or data dependency error for A. MATLAB reports only structural issues that prevent further analysis.

2. Resolve nested parfor-loop errors.

   You cannot nest a parfor-loop inside another parfor-loop. Choose one loop to run in parallel and rewrite the other loop as a for loop. Because parallel overhead occurs only once, run the outer loop in parallel.

   parfor x = 0:0.1:1
       for y = 2:10
           A(y) = A(y-1) + y;
       end
   end
   

   Alternatively, you can call a function that uses a parfor-loop inside the body of the other parfor-loop. However, such nested parfor-loops give you no computational benefit, because all workers are used to parallelize the outermost loop. For help dealing with nested loops, see Nested parfor and for-Loops and Other parfor Requirements.

3. Fix invalid parfor-loop variables.

   In the original example, the loop variable x does not use consecutive integers, which parfor requires. Replace the loop variable with a valid integer index and map it to the required values. For next steps in troubleshooting parfor-loop variables, see Ensure That parfor-Loop Variables Are Consecutive Integers.

   xValues = 0:0.1:1;
   parfor idx = 1:numel(xValues)
       x = xValues(idx);
       for y = 2:10
           A(y) = A(y-1) + y;
       end
   end
   

4. Diagnose loop independence and variable dependencies.

   After you fix the structural issues, MATLAB analyzes the loop body. Code Analyzer now reports that variable A cannot be classified.

   

   The body of the parfor-loop must be independent. One loop iteration cannot depend on a previous iteration, because the iterations are executed in parallel in a nondeterministic order. In the example,

   A(y) = A(y-1) + y;

   is not independent, and therefore you cannot use parfor. To proceed, restructure the computation to remove the dependency or keep the loop as a for-loop. For next steps in dealing with independence issues, see Ensure That parfor-Loop Iterations Are Independent.