How to Keep Rows and Columns headers when applying matrix Operation in matlab?

dear experienced:

I have dataset stored in matlab, which contain 100+ rows (instances) and 20 columns (features), i have read data used

A= xlsread(filename);

when i need to apply such operation like :

 Anorm =normc(A) [ normalized matrix of original matrix A]

then when i write normalized matrix, i would to keep the rows and columns headers as its in original matrix, following image just show an example used to explain my instances names and variable names.

My question: how we can apply such previous functions and keep rows and columns headers of matrix?

example of input .. and output matrix ..

Answer 1

[Image Analyst]

That is not a dataset variable. num is a double matrix, and txt and raw are cell arrays.

You can transpose any of them. But what does it mean to compute the distance transform of a cell array of strings? Or of a cell array?

[ahmed obaid]

i have updated my question, thanks

[Image Analyst]

You can say

[num, txt, raw] = xlsread(filename);
% Scale numbers:
scaledNumbers = mat2gray(num); % Scale min to 0, max to 1
% Replace numbers in raw with scaled version.
[rows, columns] = size(raw);
for col = 1 : columns
    for row = 1 : rows
        raw{row, col} = scaledNumbers(row, col);
    end
end

[ahmed obaid]

Index exceeds matrix dimensions, is there any issue in code or matrix please?

[Image Analyst]

Try this:

[num, txt, raw] = xlsread(filename);
% Scale numbers:
scaledNumbers = mat2gray(num); % Scale min to 0, max to 1
% Replace numbers in raw with scaled version.
[rows, columns] = size(raw);
for col = 2 : columns
    for row = 2 : rows
        raw{row, col} = scaledNumbers(row-1, col-1);
    end
end

[ahmed obaid]

excellent, its work very fine .. only one doubt please: is this code normalize vectors in independent way? in other words max value is calculated over all matrix or independent for each vector ?

[Image Analyst]

It looks over the entire 2-D matrix and finds the min and max, then sets the min to 0, the max to 1, and everything in between is scaled linearly. It does not find minima and maxima on a row-by-row or column-by-column basis. For that, you'd have to call mat2gray with just one row or column and loop over rows or columns.