% KNNCLASS k-Nearest Neighbours classifier.
%
% Synopsis:
%  y = knnclass(X,model)
%
% Description:
%  The input feature vectors X are classified using the K-NN
%  rule defined by the input model.
% 
% Input:
%  X [dim x num_data] Data to be classified.
%  model [struct] Model of K-NN classfier:
%   .X [dim x num_prototypes] Prototypes.
%   .y [1 x num_prototypes] Labels of prototypes.
%   .K [1x1] Number of used nearest-neighbours.
%
% Output:
%  y [1 x num_data] Classified labels of testing data.
%
% Example:
  trn = load('riply_trn');
  subplot(2,2,1);title('訓練數據集分布');
  % 原始分布
   ppatterns( trn );
   
  tst = load('riply_tst');
  subplot(2,2,2);title('測試數據集分布');
  % 原始分布
   ppatterns( tst ); 
   gauss_model = mlcgmm(trn);%1
  
  t=cputime;   %1              %計時開始
  
 quad_model = bayesdf(gauss_model);%1 
  ypred = knnclass(tst.X,knnrule(trn,5));
  cerror( ypred, tst.y )
  
  text1=num2str(error);     % 計算分類誤差
   
  subplot(2,2,3);title('訓練數據集pgauss分布');
  

  ppatterns(trn); pboundary(quad_model);  %1
 
 
  subplot(2,2,4);title('測試數據集pgmm分布');
  ppatterns( tst );%pgmm( model ); 
  pboundary(quad_model);  %1
  %2 pboundary(quad_model); 
  
  hpop1 = uicontrol('Style', 'text',...
       'String','分類誤差', 'Position', [100 -20 100 50], 'FontSize',16);
  hpop2 = uicontrol('Style', 'text',...
       'String', text1, 'Position', [200 -20 100 50], 'FontSize',16);
  hpop3 = uicontrol('Style', 'text',...
       'String','時間耗費', 'Position', [300 -20 100 50], 'FontSize',16);
  hpop4 = uicontrol('Style', 'text',...
       'String', TimeCost, 'Position', [400 -20 100 50], 'FontSize',16);