MATLAB: Classification error: slicing struct for parfor loop

classificationparallel computingparforslicing

I'm trying to run several calibrations in parallel, however even after slicing all the variables I still get the error: The variable parforloopData in a parfor cannot be classified. Here's the abbreviated code:
   parfor (i=1:size(uniqueDates,1))
       if( parforloopData(i,1).skip == 0 )
        while (parforloopData(i,1).gbest > DEParams.VTR)
            % start differential evolution
            [parforloopData(i,1).bestmem, parforloopData(i,1).bestval, parforloopData(i,1).bestFctParams,   parforloopData(i,1).iterno, parforloopData(i,1).pop] =  differentialevolution(  DEParams, paramDefCell, objFctHandle, parforloopData(i,1), parforloopData(i,1).objFctParams,  emailParams, optimInfo); 
            parforloopData(i,1).gbest = parforloopData(i,1).bestval;
            for(k=1:size(parforloopData(i,1).pop,1))
                parforloopData(i,1).ObjectiveFunction = @(x) SVJObjFunDEGulli(parforloopData(i,1), x);
                [parforloopData(i,1).posbestSet, parforloopData(i,1).posgbest,parforloopData(i,1).exitflag,  parforloopData(i,1).output] = fminsearch(parforloopData(i,1).ObjectiveFunction, parforloopData(i,1).pop(k,:));
                if (parforloopData(i,1).posgbest < parforloopData(i,1).gbest)
                    parforloopData(i,1).bestSet = parforloopData(i,1).posbestSet;
                    parforloopData(i,1).gbest = parforloopData(i,1).posgbest;
                end
            end
            if ( abs(parforloopData(i,1).bestLocalMinimum - parforloopData(i,1).gbest) > reseterrgrad)
                parforloopData(i,1).bestLocalMinimum = parforloopData(i,1).gbest
            else
                parforloopData(i,1).newSet = minbound+(maxbound-minbound).*rand(1,8);
                while (~checkStartingValuesDifference(parforloopData(i,1).newSet,parforloopData(i,1).startingValuesMatrix, diffInStartingValues ))
                    parforloopData(i,1).newSet = minbound+(maxbound-minbound).*rand(1,8);
                end
                parforloopData(i,1).startingValuesMatrix(end+1,:) = parforloopData(i,1).newSet;
                parforloopData(i,1).objFctParams.V0 =  parforloopData(i,1).newSet(1)
            end
        end
        optimParams(i,:) = [uniqueDates(i) parforloopData(i,1).bestSet parforloopData(i,1).gbest];
    end
end
parforloopData is a struct, which has many of its variables entered in a previous normal loop. I don't really understand why I get the error here, since each loop works on its own slice of parforloopData

Best Answer

It might help to store parforloopData(i,1) in a temporary variable inside the loop, use the temporary variable everywhere, and then update parforloopData(i,1) with the value of the temporary before exiting the parfor loop.
Related Question