nearest_neighbor.m

% Nearest neighbor training error rate is 0%.
% Nearest neighbor test error rate is 16%.

%This function takes in a training data matrix Xtrain, training
%label vector ytrain and uses them to compute the average cat
%and dog vectors. It also takes in a data matrix Xrun and 
%produces a vector of label guesses yguess. Each guess is found
%by searching through Xtrain to find the closest row, and then 
%outputting its label.
function yguess = nearest_neighbor(Xtrain,ytrain,Xrun)

[row, col] = size(Xrun);
[rowT, colT] = size(Xtrain);
yguess = zeros(row, 1);
distances = zeros(rowT, 1);

for i = 1:row
    for j = 1:size(Xtrain)
        distances(j) = norm(Xtrain(j,:) - Xrun(i,:));
    end
    minimum = min(min(distances));
    yguess(i) = ytrain(find(distances == minimum));
end

%Your code should go above this line.
if (~iscolumn(yguess))
    error("yguess is not a column vector.")
elseif (length(yguess)~=size(Xrun,1))
    error("Length of yguess is not equal to the number of rows in Xrun.")
elseif (sum(unique(abs(ytrain))~=1))
    warning("Some elements in ytrain are not +1 or -1.")
elseif (sum(unique(abs(yguess))~=1))
    warning("Some elements in yguess are not +1 or -1.")
end

end