MATLAB: Does the standalone matlab (executable) code including the parallel computing feature do not work

MATLABMATLAB CompilerParallel Computing Toolbox

The (compiled) standalone matlab code (.exe) with the parallel computing/multi-core processing feature do not work even though the matlab script (.m) perfectly executes the task.

Is this the Matlab R2011b software limitation? Is there any solution to this problem?

Best Answer

NEW ANSWER:

Turn your script into a function. This is the same issue seen here for Matlab 2011 version:

https://www.mathworks.com/matlabcentral/answers/19169-parfor-error-after-compiled

OLD ANSWER:

https://www.mathworks.com/support/bugreports/1251281

Seems like it's a known issue and thus you must use a workaround

gcp('nocreate')

Original Q&A here:

https://www.mathworks.com/matlabcentral/answers/228910-standalone-application-compiler-toolobox-of-a-code-using-the-parallel-computing-toolbox

Related Solutions

MATLAB: Using the Java library compiler, I converted gcp but Matlab can’t find it

We found out that the machine we were using didn't have the toolbox loaded when we thought it did, so the function never got exported

MATLAB: Parfor errors when file is compiled

This type of error occurs if you are attempting to compile a script instead of a function. Try changing your code to:

function myParforTest
  try
    if(isdeployed)
        matfile = 'local.mat'; % can also use uigetfile to let
        setmcruserdata('ParallelConfigurationFile',matfile);
    end
  if (matlabpool('size') > 0)==1 %cerrar matlabpool si está abierto
    matlabpool close
  end
  matlabpool open 2;
  parfor i=1:10
  end
  matlabpool close;
  uiwait(msgbox('Success'))
  catch ME
    uiwait(msgbox(ME.message))
  end
end

Also, make sure that the local.mat file is included in the CTF archive (-a option if using mcc).

Best Answer

Related Solutions

MATLAB: Using the Java library compiler, I converted gcp but Matlab can’t find it

MATLAB: Parfor errors when file is compiled

Related Question