clear

close all

% path to folder containing training data

old_trainset_folder = '/home/andrew/mitosis/MITOS/testing/';

% make training set directories

new_trainset_folder = '/home/andrew/mitosis/mitosis-test/';

if exist(new_trainset_folder, 'dir')

    rmdir(new_trainset_folder, 's')

end

mkdir([new_trainset_folder 'true'])

mkdir([new_trainset_folder 'false'])

% count the total number of images for the waitbar

N = 0;

for j=1:5

   % find the number of images in the folder

    folder = [old_trainset_folder 'A' num2str(j-1, '%02u') '_v2/'];

    n = length(dir([folder '*.csv']));

    N = N + n;

end

% iterate over the twelve patients

M = 0;

P = 0;

k = 0;

h = waitbar(0,'Creating training set ... 0 %');

for j=1:5

    % find the number of images in the folder

    folder = [old_trainset_folder 'A' num2str(j-1, '%02u') '_v2/'];

    image_files = dir([folder '*.bmp']);

    csv_files = dir([folder '*.csv']);

    n = length(image_files);

    % create new data set

    for i=1:n

        image_file = [folder image_files(i).name];

        csv_file = [folder csv_files(i).name];

        [m, p] = add_dataset(image_file, csv_file, new_trainset_folder);

        M = M + m;

        P = P + p;

        k = k + 1;

        waitbar(k/N,h,['Creating training set ... ' num2str(100*k/N) ' %']);

    end

end

close(h)

disp(['Created ' num2str(P) ' window training images from ' num2str(N) ' large training images containing ' num2str(M) ' mitotic figures.'])