function bestSubject = knnFace(facePath, k, dctData, dctDim, dctDataClass)
%{
    facefile    path to the face image file
    k           the number of close neighbors to find
    dctData     the pretrained dataset
    dctDim      the dimesion of DCT data to work with
%}

% Find the number of faces in the dataset
N = length(dctData);

% Get the 1-D feature vector of the DCT of the face
features = findfeatures(facePath, dctDim);

% Find the distance from the face to each face in the dataset
dists = zeros(N, 1);
for idx = 1 : N
    % Calculate the Euclidean distance in any dimension (found with 2-norm)
    dists(idx) = norm(features' - dctData(idx,:), 2);
end

% Sort the distances in ascending order
[sDists, origIdx] = sort(dists);

% Find the corresponding best subjects for the shortest distances
bestSubjects = zeros(k,1);
for idx = 1 : k
  bestSubjects(idx) = dctDataClass(origIdx(idx));
end

bestSubject = mode(bestSubjects);