normalization of cell arrays

Hi

i want to normalize the values in a cell array called G

G is partitioned into 6 blocks (cell arrays) so i want to normalize each block separately (G{i,j}), each block is of size 128X128

Gmin = min((G(:))); 

Gnorm = (G - Gmin) ./ (max(G1{:,:}) - Gmin);

i tried this:

Gmin(i,j)=min(G{i,j});

but i got this error:

??? Undefined function or method 'min' for input arguments of type 'cell'.

Error in ==> DOC at 124
        Gmin(i,j)=min(G{i,j});

any suggestions please ???

Answer 1

[Walter Roberson]

Gmin = cellfun(@(M) min(M(:)), G, 'Uniform', 0);
Gmax = cellfun(@(M) max(M(:)), G, 'Uniform', 0);

Gnorm = cellfun(@(M,minM,maxM) (M-minM) ./ (maxM-minM)), G, Gmin, Gmax, 'Uniform', 0);

This relies on two uncommon usages:

• deliberately returning a cell array when the results would fit in a vector
• passing multiple cell arrays in to cellfun(), knowing that the corresponding elements will be extracted from each and passed to the function handle

There are other ways of coding this, such as using arrayfun()

[yasmine]

unfortunately,i had this error:

??? Undefined function or method 'min' for input arguments of type 'cell'.

Error in ==> @(M)min(M(:))

Error in ==> DOC at 129

Gmin = cellfun(@(M) min(M(:)), G, 'Uniform', 0);

[Asma]

the error occurs probably due to the presence of non-numeric or absent values, try checking for isnumeric before proceeding

% Check which cells are numeric isNumericCell = cellfun(@isnumeric, G);