how to storage output data from a nested parfor loop?

Question

Beatriz Sanchez 2018-6-28

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/407773-how-to-storage-output-data-from-a-nested-parfor-loop

编辑： Beatriz Sanchez 2018-7-3

Hi, so I have something like this:

m=[150; 300; 0]
m(1)=180
vec=[-180 -36 0 36 180]
d=1000
for i=1:5
    dm2=vec(:,i)
    m(2)=dm2+m(1)
    for ind=1:5
        dm3=vec(:,ind)
        m(3)=dm3+m(1)
        parfor k=1:5
            a=randi (1000,3,1)
            out=floor(a/sum(a)*d)
            out(1)=out(1)-sum(out)+d
            x= out .* m
            x'
        end
    end
end

But obviously, the value of x' it's replaced on each iteration, how can I storage this data? I was thinking on storage it in a multidimensional array like x(25,3,5).

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Edric Ellis 2018-6-29

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/407773-how-to-storage-output-data-from-a-nested-parfor-loop#answer_326806

在 MATLAB Online 中打开

In this case, x is a 3-element vector. You can store each of these vectors in a 4-D array like so:

for i = 1:5
  ...
  for ind = 1:5
    ...
    parfor k = 1:5
      ...
      output(i, ind, k, :) = x;
    end
  end
end

2 个评论
显示无隐藏无

Beatriz Sanchez 2018-7-3

Thank you very much for your help. Unfortunately it gives me an error, it says:

'Error using pueba>(parfor consume)

Assignment has more non-singleton rhs dimensions than non-singleton subscripts'

Error in pueba (line 14) parfor k= 1:5'

Beatriz Sanchez 2018-7-3

编辑：Beatriz Sanchez 2018-7-3

在 MATLAB Online 中打开

alright, I stablish F before the loop like this:

F=zeros(5,5,5,3)
for i=5
...
for ind=5
...
parfor k=5
...
F(i,ind,k,:)=X'
end 
end
end

and the code runs perfectly, only that the 'k' (or third) dimension appers with a random order, but that is not important for what I want, and I figure that's because the parfor loop run independently from each iteration.

Anyways, I appreciate a lot your help, thank you :-)

请先登录，再进行评论。