Clear Filters
Clear Filters

Extracting data from struct as matrix

14 views (last 30 days)
Hi,
I have the following struct:
S(1).a = 1:5;
S(2).a = 11:10;
S(3).a = 21:27;
Is there a way to vertically concatenate them without looping? I cannot do vertcat(S.a) because data is of different size.
I want output like this [1:5; 11:10; 21:27]
Thank you
  2 Comments
Ganesh
Ganesh on 13 Jun 2024
A matrix with these dimensions cannot be constructed.
a = [1:5; 11:10; 21:27]
Error using vertcat
Dimensions of arrays being concatenated are not consistent.
Stephen23
Stephen23 on 13 Jun 2024
"I want output like this [1:5; 11:10; 21:27]"
Matrices cannot have rows of different lengths.

Sign in to comment.

Accepted Answer

Ganesh
Ganesh on 13 Jun 2024
You can consider the following workaround for your problem, without using for loops:
S(1).a = 1:10;
S(2).a = 11:17;
S(3).a = 21:26;
aCell = arrayfun(@(x) {x.a}, S);
% Find the maximum length among all arrays
maxLength = max(cellfun(@numel, aCell));
% Function to pad arrays with NaN to match the maxLength
padFunction = @(x) [x, NaN(1, maxLength - numel(x))];
y = cellfun(padFunction, aCell, 'UniformOutput', false);
y = vertcat(y{:})
y = 3x10
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 NaN NaN NaN 21 22 23 24 25 26 NaN NaN NaN NaN
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
  2 Comments
Rashi Monga
Rashi Monga on 13 Jun 2024
Thanks everyone for their replies. It was helpful.
Stephen23
Stephen23 on 14 Jun 2024
Alternatively doanload PADCAT here:
and use it like this:
S(1).a = 1:5;
S(2).a = 10:11;
S(3).a = 21:27;
M = padcat(S.a)
M = 3x7
1 2 3 4 5 NaN NaN 10 11 NaN NaN NaN NaN NaN 21 22 23 24 25 26 27
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>

Sign in to comment.

More Answers (2)

Jonas
Jonas on 13 Jun 2024
Edited: Jonas on 13 Jun 2024
as you already said, you canot concatenate the data as matrix, since it has different size. you could use a cell array, but this would be no better than the struct. another possibility coul be to pad shorter variables with NaN. but here, you would need a loop.
S(1).a = 1:5;
S(2).a = 11:10;
S(3).a = 21:27;
sizes=arrayfun(@(in) length(in.a),S);
maxSize=max(sizes);
asMatrix=nan(numel(sizes),maxSize);
for sNr=1:numel(S)
asMatrix(sNr,1:sizes(sNr))=S(sNr).a;
end
asMatrix
asMatrix = 3x7
1 2 3 4 5 NaN NaN NaN NaN NaN NaN NaN NaN NaN 21 22 23 24 25 26 27
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>

Shivani
Shivani on 13 Jun 2024
Edited: Shivani on 13 Jun 2024
You will need to modify the solution in the above answer as seen in the below code snippet to obtain the desired result.
S(1).a = 1:5;
S(2).a = 11:15;
S(3).a = 21:25;
M = vertcat(S.a)
M = 3x5
1 2 3 4 5 11 12 13 14 15 21 22 23 24 25
<mw-icon class=""></mw-icon>
<mw-icon class=""></mw-icon>
Please note that the ranges you have provided cannot be concatenated vertically because they will not be of the same length. I am therefore assuming the ranges to be 1:5, 11:15 and 21:25. Since they all contain 5 elements, we can concatenate them to a matrix containing 5 columns.
Hope this helps!

Tags

Community Treasure Hunt

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

Start Hunting!