function  [Labels, DecisionValue]= SVMClass(Samples,AlphaY, SVs, Bias, Parameters)
% USAGE:
% [Labels, DecisionValue]= SVMClass(Samples,AlphaY, SVs, Bias, Parameters)
%
% DESCRIPTION:
% Classify a group of input patterns given a trained svm classifier.
% In fact, this function is used to do the input parameter checking, and it 
% depends on a mex file to implement the algorithm.
% 
% INPUTS:
% Samples: all the input patterns. (a row of column vectors)
% AlphaY: Alpha * Y, where Alpha is the non-zero Lagrange Coefficients
%                    Y is the corresponding labels.(a row vector)
% SVs : support vectors. That is, the patterns corresponding the non-zero
%       Alphas.(a row of column vectors)
% Bias : the bias in the decision function, which is AlphaY*Kernel(SVs',x)-Bias. (scalar)
% Parameters: the paramters required by the training algorithm. 
%             (a 10-element row vector)
%            +-----------------------------------------------------------------
%            |Kernel Type| Degree | Gamma | Coefficient | C |Cache size|epsilon| 
%            +-----------------------------------------------------------------
%				 -------------------------------------------+
%            | SVM Type |	nu (nu-svm) | loss tolerance |				
%				 -------------------------------------------+
%            where Kernel Type:
%                   0 --- Linear
%                   1 --- Polynomial: (Gamma*<X(:,i),X(:,j)>+Coefficient)^Degree
%                   2 --- RBF: (exp(-Gamma*|X(:,i)-X(:,j)|^2))
%                   3 --- Sigmoid: tanh(Gamma*<X(:,i),X(:,j)>+Coefficient)
%                  Gamma: If the input value is zero, Gamma will be set defautly as
%                         1/(max_pattern_dimension) in the function. If the input
%                         value is non-zero, Gamma will remain unchanged in the 
%                         function.
%                  C: Cost of the constrain violation (for C-SVC & C-SVR)
%                  Cache Size: as the buffer to hold the <X(:,i),X(:,j)> (in MB)
%                  epsilon: tolerance of termination criterion
%                  SVM Type: 
%                   0 --- c-SVM classifier
%                   1 --- nu-SVM classifier
%                   2 --- 1-SVM 
%                   3 --- c-SVM regressioner 
%                  nu: the nu used in nu-SVM classifer (for 1-SVM and nu-SVM)
%						 loss tolerance: the epsilon in epsilon-insensitive loss function	
%
% OUTPUTS:
% ClassRate: the ration of properly classified pattern to the total number of
%            input patterns.
% DecisionValue: the output of the decision function.
%
if (nargin ~= 5)
   disp(' Incorrect number of input variables.');
   help SVMClass;
   return
end
[spM spN]=size(Samples);

[prM prN]= size(Parameters);
if prM ~= 1
   disp(' Error: ''Parameters'' should be a row vector.');
   return
end

if prN ~= 10
   disp(' Error: ''Parameters'' should have exactly 10 elements.');
   return
end

if (Parameters(1)>3) & (Parameters(1) < 0)
   disp(' Error: this program only supports 4 types of kernel functions.');
   return
end

if (Parameters(8)>3) & (Parameters(8) < 0)
   disp(' Error: this program only supports 4 types of SVMs.');
   return
end

[svM svN]=size(SVs);
[ayM ayN]=size(AlphaY);
if ayM ~= 1
   disp(' Error: ''AlphaY'' should be a row vector.');
   return
end

if svM ~= spM
   disp(' Warning: ''SVs'' should have the same feature dimension as ''Samples''.');
   [Samples]=DimFit(Samples,svM);
end

Labels = ones(1,spN); % this is a set of phony labels to feed the mex file.
% call the mex file
[ClassRate, DecisionValue]= mexSVMClass(Samples, Labels, AlphaY, SVs, Bias, Parameters);
% Labels are 1 or -1
Labels = sign(DecisionValue);