Make a sript faster

Hello everyone,

Does anyone know how to make this script faster knowing that I have 800 matrix to import?

pathname = uigetdir();
 
cycle = cell(2,400);
 
for i = 1:2
    for j = 1:100
        for k = 1:4
 
            filename = sprintf('Cycle%d_Image%d_%d.csv', j, k, i);
 
            if exist(fullfile(pathname, filename), 'file')~=2
                error('File not found: %s', fullfile(pathname, filename))
            end
            
            strrepInFile(fullfile(pathname, filename), ',', '.');
 
            cycle{i,4*(j-1)+k} = dlmread(fullfile(pathname, ['new_' filename]), ';');
 
            delete(fullfile(pathname, ['new_' filename]))
 
        end
    end
end

[Leeee]

Here is a part of the data and the time that the script Thank you in advance!

[Geoff Hayes]

Do you always need to call strrepInFile everytime your run the above code or could you just do it once and keep those updated files (rather than deleting them)?

[Leeee]

Hello Geoff

Yes I thought it was the right method for all of my 800 files and I do not see how to do it once.

[Leeee]

It's true it takes a lot of time the strrepInFile and suddenly I replaced it with strrep which takes less time. And how to reduce the dlmread time which represents 80% of the execution time?

[John D'Errico]

Now that you have shown the bottlenect to be file read time, you need to accept that file read time is something you have little direct control over.

If you want it to be faster, then you need to get a faster computer, or you need a disk that allows faster access time. A good choice there is a SSD drive. If you are reading files from a drive that is not local, then move them to be local FIRST. Then read them in from a SSD.

[Walter Roberson]

dlmread calls textscan. You can avoid some overhead by calling textscan yourself especially if the files have fixed format.

[Leeee]

@John and @Walter,

Thank you for your answers I admit that it is super effective your contributions I am delighted. Thanks again!