Pad with Nan according to maximum row in cell

Hi, I have a 1 x 8 cell (attached) and would like to cell2mat this matrix i.e to expand and show into one big numeric 97 x 8 array (97 being the maximum length of a cell array). I would also like to pad the ends of the smaller cell arrays with Nan to make all columns the same row length. How may I write for this?

Thank you!

[Stephen23]

"How may I write for this?"

Rather than reinventing the wheel, you can simply download Jos' excellent PADCAT here:

https://www.mathworks.com/matlabcentral/fileexchange/22909-padcat

and use it like this, where C is your cell array:

M = padcat(C{:})

This works automatically for vectors in either orientation.

Answer 1

[Fabio Freschi]

I keep all steps separated

% the data
A = {rand(47,1),rand(80,1),rand(97,1)};

% pad with NaNs
A = cellfun(@(x)[x(1:end); NaN(97-length(x),1)],A,'UniformOutput',false);
 
% make a matrix
A = cell2mat(A);

[Joanne Hall]

Hi Fabio,

How do I switch the dimensions of the data in your first line of code, so that, for example, rand(47,1) is then switched to rand(1,47), and then run/execute the rest of your helpful code?

(I'm new to cell arrays and I have the same problem, except the dimensions in my cell array are switched, ex. 1X138702).

Thanks in advance!

Joanne

[Fabio Freschi]

Joanne,

if I understand correctly your question, you must simply change everything that is row into column and vice versa

% the data
A = {rand(1,47);rand(1,80);rand(1,97)};  % <- note the semicolon, note the row/col excange in the data
% pad with NaNs
A = cellfun(@(x)[x(1:end) NaN(1,97-length(x))],A,'UniformOutput',false);  % <- note the missing semicolon after x(1:end) and the row/col excange in NaN
 
% make a matrix
A = cell2mat(A);

Hope iot helps

[Tobias]

Hi! Thank you for your answer, unfortunately i appears that for cells < max_t the function just adds another cell arrayto the end of it, with max_t size. Do you know how to circumvent this ?