Info

This question is closed. Reopen it to edit or answer.

How to create doubles from cell arrays?

1 view (last 30 days)
gsourop
gsourop on 27 Apr 2018
Closed: MATLAB Answer Bot on 20 Aug 2021
I would like to ask if there is a more efficient code to do the following task:
a = cell(10,1);
for i = 1 : 10
a{i,1} = randn(200,5);
end
for j =1:5
b{j} = [a{1,1}(:,j) a{2,1}(:,j) a{3,1}(:,j) a{4,1}(:,j) a{5,1}(:,j)];
end
Thank you!

Answers (1)

Guillaume
Guillaume on 27 Apr 2018
Edited: Guillaume on 27 Apr 2018
if there is a more efficient code
Yes, don't use cell arrays!
a = randn(200, 5, 10);
your b{j} is simply
permute(a(:, j, :), [1 3 2])
%or
squeeze(a(:, j, :))
Even simpler, move the j to the third dimension
a = randn(200, 10, 5);
%b{j} would simply be a(:, :, j)
  2 Comments
gsourop
gsourop on 27 Apr 2018
The initial matrices are of this form, I just use randn values in order to replicate the experiment. It is not that I am now starting the calculations. So I have to achieve result 'b' given the form of 'a'.
Guillaume
Guillaume on 27 Apr 2018
Yes, I also used randn purely for demonstration.
Don't use cell arrays if you don't have to. It complicates everything. Since the matrices in your cell array are all the same concatenate them into a 3D matrix and use that from then on.
%a: a cell array of MxN matrices;
bettera = permute(cat(3, a{:}), [1 3 2])
From then on, whenever you wanted to use b{j}, use
bettera(:, :, j)
It is even possible that whatever you wanted to do with each b{j] could be done at once on bettera.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!