parallel_function error, "V must be a 3-D array"

Patrick

2012 3 9

0 个回答

2 次查看（30 天）

0 个投票

Hi all, I'm trying to parallelize some of my code, and I have no idea how to fix this error. Here's the code:

energy=zeros(length(epsilon_theta),length(epsilon_r));
Emag=zeros(length(epsilon_theta),length(epsilon_r),21);
parfor m=1:length(epsilon_theta)
    energy1=zeros(length(epsilon_r));
    Emag1=zeros(length(epsilon_r),21);
    for n=1:length(epsilon_r)
      [energyf, Emagf, Source_Emag]=DDA_ExpRf_fun([epsilon_theta(m) epsilon_r(n)]);
      energy1(n)=energyf;
      Emag1(n,:)=Emagf;
    end
    energy(m,:)=energy1;
    Emag(m,:,:)=Emag1;
end

When I run it, parfor breaks and returns

??? Error using ==> parallel_function at 598 Error in ==> interp3>linear at 320 V must be a 3-D array.

Does anyone know what could be the problem? Thanks! I'm totally new to parallel computing. Any other general tips on optimizing would be great as well.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Edric Ellis 2012-3-9

Do thinks work correctly with MATLABPOOL closed? Do they work correctly with a FOR loop instead of a PARFOR loop? I presume your code in DDA_ExpRf_fun is calling interp3 - perhaps you should add some error checking that your variables are the correct size when you get there.

请先登录，再进行评论。

请先登录，再回答此问题。

Follow Question