% outputs a graphical representation of the clustering solution
function view_clusters(points, centroids)
    % TODO graphical representation coded here

  % number of points
  NP = rows(points);
  % number of clusters
  NC = rows(centroids);

  % function that leads the interval (0, 1) in the interval ((i - 1) / NC, i / NC)
  get_color = @(i) (rand(1 ,1) + i - 1) / NC;
  color = arrayfun(get_color, [randperm(NC); randperm(NC); randperm(NC)]);

  if NC == 1
    scatter3(points(:, 3), points(:, 2), points(:, 1),
             [10], [rand(1, 1) rand(1, 1) rand(1, 1)], 'filled');
    hold on;
    return
  endif

  % function that get for each column of x the row where is the min
  f = @(x) [i j] = find(x == min(x));

  % the cluster of each point
  cluster = f(reshape(norm(repmat(centroids, NP, 1) - kron(points, ones(NC, 1)), 'rows'), NC, NP));

  for i = 1 : NC
    v = points(find(cluster == i), :);

    scatter3(v(:, 3), v(:, 2), v(:, 1),
             [10], [color(i, 1) color(i, 2) color(i, 3)], 'filled');
    hold on;
  end

end