Get Rank-Maximal Allocation
2 views (last 30 days)
Show older comments
Input: n subjects ranked i items (in example below i = 3) by preference, e.g.:
% Rank: 1 2 3
Alice = [x y z];
Bobby = [y x z];
Carlo = [x z y];
S = [Alice; Bobby; Carlo];
Problem: Find a fair allocation of items to subjects, e.g. through Rank-maximal allocation (but not necessarily)
Desired output:
% Subjects[A; B; C]
outRank = [1; 1; 2]; % Tells me the rank of the received items
% OR:
outItem = [x; y; z]; % Tells me directly which subject gets which item. outItem = S(i,outRank(i))
Wikipedia article about the problem with algorithm. I just can't make sense of the outputs of any of the MATLAB functions with seemingly useful names.
0 Comments
Answers (1)
Anurag
on 25 Oct 2023
Hi Stefan,
I understand that you want to implement the Rank-Maximal allocation algorithm, please refer to the following code for the same.
% Input: n subjects ranked i items
% Create a matrix where each row represents a subject's preference ranking
Alice = [2, 1, 3];
Bobby = [1, 2, 3];
Carlo = [2, 3, 1];
% Stack the preferences of all subjects into a single matrix
S = [Alice; Bobby; Carlo];
% Number of subjects and items
[n, i] = size(S);
% Initialize output arrays
outRank = zeros(n, 1);
outItem = strings(n, 1); % Initialize outItem as a string array
% Find the rank-maximal allocation
for subject = 1:n
% Find the item with the lowest rank for the current subject
[~, minRank] = min(S(subject, :));
% Find the subject who prefers this item the most
[~, maxPreference] = max(S(:, minRank));
% Assign the item with the lowest rank to the subject who prefers it the most
outRank(maxPreference) = minRank;
outItem(maxPreference) = char('A' + subject); % Assign the subject letter
% Set the rank of the assigned item to a high value to avoid reassignment
S(maxPreference, minRank) = i + 1;
end
% Output the results
disp('Subjects [A; B; C]:');
disp(['outRank = ', num2str(outRank')]);
disp(['outItem = ', outItem']);
Hope this helped.
Regards,
Anurag
0 Comments
See Also
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!