Extracting values out of a cell array with tables

I have a 41x1 cell array with tables including stock returns (find the cell array attached). The tables dimensions vary in number of columns.

I would now like to exctract all the data except column 1:3 from every table into one 1xXXX matrix.

Does anyone have a suggestion for a loop working throug my cell array that returns the described matrix?

Answer 1

[per isakson]

This script puts all the numerical data into a row vector (column-wise). You didn't say anything about order.

%%
load('CA.mat')
%%
XXX = [];
for jj = 1 : numel( DL12_10T )
    tbl = DL12_10T{jj}(:,4:end); 
    num = table2array( tbl ); 
    XXX = horzcat( XXX, reshape(num,1,[]) );
end
whos XXX

displays

  Name      Size                  Bytes  Class     Attributes
  XXX       1x301836            2414688  double     

[Fabian Niederreiter]

Returns the data excactly in the format I desired. Thank you!

There's no way to easily preallocate XXX with the right size is it? Cause in Matlab I get the warning that XXX should be preallocated due to performance reasons.

[per isakson]

It's possible to preallocate, but if the script is fast enough, why bother? It's easy to suppress the warning, which shows that you made a decision. If nothing else, preallocation adds a bit to the complexity of the code.

In this case you could preallocate a large XXX, and at the end remove any excess elements:

sas = whos('DL12_10T');
large_length = sas.bytes/8;     % assuming double 
XXX = nan( 1, large_length );   % physical memory shouldn't be a problem

See Preallocation performance

[Fabian Niederreiter]

Thanks for the information on the possiblity to surpress the warning. That was totally reasonable for my cause. :)

Thanks again for your time and effort! I appreciate it.