How to use cell2mat for specific elements of cell array?

20 views (last 30 days)
Hi everyone,
I am trying to create a double from specific elements of a cell array. If I have a cell array B, of dimensions 6x5, of which each element contains a double 10x4, such as:
for i = 1 : 6
for j = 1 : 5
A = randn(10,4)
B{i,j} = A;
I would like to create a big double out of all the rows and first 3 columns of B, and the second and third column of each interior double A (on this 6x3 cell array). Hence, the expected double C that should be a 60x6 matrix. I am using
C = cell2mat(B{:,1:3}(:,[2,3]));
But I get an error. What am I missing here?

Accepted Answer

Stephen23 on 25 Oct 2018
Edited: Stephen23 on 25 Oct 2018
C = cell2mat(B(:,1:3));
C = C(:,2:3);
"What am I missing here?"
Several things: firstly, that the B{...} syntax creates a comma-separated list from the contents of the cell array, which in general is not a valid input for cell2mat:
Secondly, that in MATLAB indexing cannot be arbitrarily added on to the end of any other expression, so you will need to use a temporary variable.
Stephen23 on 25 Oct 2018
Edited: Stephen23 on 25 Oct 2018
"Is it an efficient way to proceed?"
Not the way you are doing it, because you expand the array D on each loop iteration, which forces MATLAB to move it in memory. You should preallocate the output array and use indexing:
But ultimately you will just have to test it and find out.
Personally I would not use a loop, simply because it pointlessly obfuscates what you are doing: the time wasted trying to read/understand/maintain/reverse-engineer that loop will likely be much more than any possible gain you might get from runtime efficiency (or not). Unless this really is the slowest part of your code, you should write using the clearest syntax possible (i.e. without a loop, e.g. my answer). Why bother writing nine lines of code just to do the same thing as my two lines? This is not "efficient" coding.

Sign in to comment.

More Answers (0)


Find more on Data Type Conversion in Help Center and File Exchange


Community Treasure Hunt

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

Start Hunting!