normalise.m.svn-base

a function inside machine learning

function [normalisedX1, normalisedX2] = normalise(X1, X2)
% Normalise the features (columns) of matrices X1 (and optionally X2) such that 
% each feature of X1 has unit norm. X1 and X2 have examples as their rows. 
%
% Usage: [normalisedX1, normalisedX2] = normalise(X1, X2)
% Inputs/Outputs: 
%   X1 - an (l x n) matrix whose rows are examples
%   X2 (optional) - an (l2 x m) matrix whose rows are examples
%
%   normalisedX1 - normalised X1 
%   normalisedX2 (optional) - normalised X2 
%
% Copyright (C) 2006 Charanpal Dhanjal 

% This library is free software; you can redistribute it and/or
% modify it under the terms of the GNU Lesser General Public
% License as published by the Free Software Foundation; either
% version 2.1 of the License, or (at your option) any later version.
% 
% This library is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
% Lesser General Public License for more details.
% 
% You should have received a copy of the GNU Lesser General Public
% License along with this library; if not, write to the Free Software
% Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA

if (nargin < 1)
    fprintf('%s\n', help(sprintf('%s', mfilename)));
    error('Incorrect number of inputs - see above usage instructions.');
end

numFeatures = size(X1, 2);

%Just make sure the training example features have unit norm 
featureNorms = sqrt(sum(X1.^2));

%Bit of cheat to make sure we don't divide by zero
zeros = featureNorms == 0;
featureNorms = 1./(featureNorms+zeros);

if ~issparse(X1)
    diagNorms = diag(featureNorms); 
else    
    diagNorms = speye(numFeatures); 

    for i=1:numFeatures
        diagNorms(i, i) = featureNorms(i);
    end 
end

normalisedX1 = X1*diagNorms;

if (nargin == 2)
    normalisedX2 = X2*diagNorms; 
end