function model=svmlight(data,options)% SVMLIGHT Interface to SVM^{light} software.%% Synopsis:%  model = svmlight(data)%  model = svmlight(data,options)%% Description:%  This function serves as an interface between Matlab %  and SVM^{light} (Version: 5.00) optimizer which trains %  the Support Vector Machines classifier.%%  The executable file 'svm_learn' must be in the path. %  The SVM^{light} software can be downloaded from:%      http://svmlight.joachims.org/%%  This function creates temporary files 'tmp_alphaXX.txt', %  'tmp_examplesXX.txt', 'tmp_modelXX.txt' and 'tmp_verbXX.txt' for %  comunication with the SVM^{light} software. The XX=datestr(now)%  is string consisting of current date and time.%           % Input:%  data [struct] Labeled binary data:%   .X [dim x num_data] Training vectors.%   .y [1 x num_data] Labels of training data (1 or 2).%  %  options [struct] Control parameters:%   .ker [string] Kernel identifier: %      'linear' (default),'rbf' and 'poly'. %   .arg [1x1] Kernel argument (default []).%   .C [1x1] SVM regularization constant (default C=inf).%   .j [1x1] Cost-factor, by which training errors on %     positive examples outweight errors on negative examples (default 1).%   .eps [1x1] Tolerance of KKT-conditions (default eps=0.001).%   .keep_files [1x1] If ==1 then keeps temporary files otherwise%     erase them.%% Output:%  model [struct] Binary SVM classifier:%   .Alpha [nsv x 1] Weights of support vectors.%   .b [1x1] Bias of decision function.%   .sv.X [dim x nsv] Support vectors.%   .sv.inx [1 x nsv] Indices of SVs (model.sv.X = data.X(:,inx)).%   .nsv [int] Number of Support Vectors.%   .kercnt [int] Number of kernel evaluations used by the SVM^{light}.%   .trnerr [real] Classification error on training data.%   .margin [real] Margin of found classifier.%   .options [struct] Copy of used options.%   .cputime [real] Used CPU time in seconds.%% Example:%  data=load('riply_trn');  %  model=svmlight(data,struct('ker','rbf','C',10,'arg',1))%  figure; ppatterns(data); psvm(model);%% See also %  SVMCLASS, XY2SVMLIGHT.%% About: Statistical Pattern Recognition Toolbox% (C) 1999-2003, Written by Vojtech Franc and Vaclav Hlavac% <a href="http://www.cvut.cz">Czech Technical University Prague</a>% <a href="http://www.feld.cvut.cz">Faculty of Electrical Engineering</a>% <a href="http://cmp.felk.cvut.cz">Center for Machine Perception</a>% Modifications:% 16-may-2004, VF% 15-jan-2004, VF, handling argument of poly kernel repared% 10-oct-2003, VF, computation of lin model added% 29-aug-2003, VF, seconds are added to the name of temporary files% 12-may-2003, VF, 1st 3 lines of verb_file are skiped% 31-jan-2003, VF, added option 'j' % 28-Jan-2003, VF% 20-jan-2003, VF, temporary files are unique and are deleted at the end% 14-Jan-2003, VF% 26-sep-2002, VF% 3-Jun-2002, V.Franctic;data=c2s(data);% gets current date and timedate_str=datestr(now);date_str(findstr(date_str,' ')) = '-';sec=clock;date_str = [date_str '-' num2str(sec(end))];% names of temporary files examples_file = ['tmp_examples' date_str '.txt'];model_file = ['tmp_model' date_str '.txt'];verb_file = ['tmp_verb' date_str '.txt'];alpha_file = ['tmp_alpha' date_str '.txt'];% make modelmodel.name = 'SVM classifier';% -- Process input arguments --------------------------if nargin < 2, options = []; else options=c2s(options); endif ~isfield(options,'ker'), options.ker = 'linear'; endif ~isfield(options,'arg'), options.arg = '[]'; endif ~isfield(options,'C'), options.C = inf; endif ~isfield(options,'eps'), options.eps = 0.001; endif ~isfield(options,'keep_files'), options.keep_files = 0; endif ~isfield(options,'j'), options.j = 1; end% gets data dimensions[dim,num_data ] = size(data);%--------------------------------switch options.ker  case 'linear'    ker='-t 0';  case 'rbf'    ker=['-t 2 -g ' num2str(1/(2*options.arg^2))];   case 'poly'     if length(options.arg) == 1,      ker=['-t 1 -r 1 -s 1 -d ' num2str(options.arg)];      else      ker=['-t 1 -s 1 -r ' num2str(options.arg(2)) ' -d ' ...           num2str(options.arg(1))];      endendcommand=['svm_learn ' ...         '-c ' num2str(options.C) ' '...         ker ' ' ...         '-v 1 ' ...         '-m 40' ' ' ...         '-j ' num2str(options.j) ' '...         '-e ' num2str(options.eps) ' '...         '-a ' alpha_file ' ' examples_file ' ' model_file ' > ' verb_file];   % converts data to SVM^light formatxy2svmlight(data,examples_file);    % call svm_learn[a,b]=unix(command);   % parses model filecheckfile(model_file); [lines]=textread(model_file,'%s');for i=1:size(lines,1)   if strcmpi( lines(i), 'threshold' )==1,     model.b=-str2num( lines{i-2});     break;   endend    checkfile(alpha_file); model.Alpha=textread(alpha_file,'%f');model.Alpha=model.Alpha(:);model.Alpha(find(data.y==2)) = -model.Alpha(find(data.y==2));checkfile(verb_file);[lines]=textread(verb_file,'%s',-1,'bufsize',5000000,'headerlines',3);for i=1:size(lines,1)  if strcmpi( lines{i}, 'misclassified,' ),    model.trnerr=str2num( lines{i-1}(2:end));    model.trnerr=model.trnerr/length(model.Alpha);  end  if strcmpi( lines(i), 'vector:' ) & strcmpi( lines(i-1), 'weight' )==1,    tmp=str2num( lines{i+1}(5:end));    if tmp~=0, model.margin=1/tmp; else model.margin=[]; end  end  if strcmpi( lines(i), 'SV:' )==1,    model.nsv=str2num( lines{i+1});  end  if strcmpi( lines(i), 'evaluations:' )==1,    model.kercnt=str2num( lines{i+1});  endendmodel.nsv = length(find(model.Alpha~=0));inx=find(model.Alpha);model.sv.X = data.X(:,inx);model.sv.y = data.y(inx);model.sv.inx = inx;model.Alpha = model.Alpha(inx);model.Alpha(find(model.sv.y==2)) = -model.Alpha(find(model.sv.y==2));if strcmp( options.ker, 'linear'),  model.W = model.sv.X * model.Alpha;endmodel.options = options;model.fun = 'svmclass';% deletes temporary filesif options.keep_files == 0,  delete(examples_file);  delete(model_file);  delete(verb_file);  delete(alpha_file);endmodel.cputime=toc;return;function checkfile(fname)% Check if file of given name exists. If not then prints% error.% attempts=0;found = exist(fname);while attempts < 5 & found  found = exist(fname);  attempts = attempts+1;  endif found == 0,  error('File %s not found.\n', fname);endreturn;%EOF