L2dist.m

% This function computes the L2 distance between the N samples 
% in the training set X and the M samples in the test set Y. 
% Samples in X and Y are arranged by rows and the resulting 
% distance matrix D is N x M where each position (i,j) 
% is the L2 distance between the i-th training sample and 
% the j-th test sample 
% function D = L2dist(X,Y)

% Efficient implementation of L2 distance without square root
% d(x,y) = \sum_d (x_d - y_d)^2 = 
%        = \sum_d x_d^2 + y_d^2 - 2*x_d*y_d
%        = \sum_d x_d^2 + \sum_d y_d^2 - 2 * \sum_d x_d*y_d
function D = L2dist(X,Y)
  XX = sum(X.*X,2);
  YY = sum(Y.*Y,2);
  D = XX + (YY' - 2*X*Y');
end

% Non-efficient implementation that computes the L2 distance for 
% each test sample with respect to all samples in the training set
% function D = L2dist(X,Y)
% for i=1:rows(Y)
%   D(:,i) = sum(abs(X-Y(i,:)).^2, 2);
%end