SVKernelMatrix.cxx

Go to the documentation of this file.

// @(#)root/tmva $Id$
// Author: Andrzej Zemla
 
/**********************************************************************************
 * Project: TMVA - a Root-integrated toolkit for multivariate data analysis       *
 * Package: TMVA                                                                  *
 * Class  : SVKernelMatrix                                                        *
 * Web    : http://tmva.sourceforge.net                                           *
 *                                                                                *
 * Description:                                                                   *
 *      Implementation                                                            *
 *                                                                                *
 * Authors (alphabetical):                                                        *
 *      Marcin Wolter  <Marcin.Wolter@cern.ch> - IFJ PAN, Krakow, Poland          *
 *      Andrzej Zemla  <azemla@cern.ch>        - IFJ PAN, Krakow, Poland          *
 *      (IFJ PAN: Henryk Niewodniczanski Inst. Nucl. Physics, Krakow, Poland)     *
 *                                                                                *
 * Minor modification to improve optimisation of kernel values:                   *
 *      Adrian Bevan   <adrian.bevan@cern.ch>  -         Queen Mary               *
 *                                                       University of London, UK *
 *      Tom Stevenson <thomas.james.stevenson@cern.ch> - Queen Mary               *
 *                                                       University of London, UK *
 *                                                                                *
 * Copyright (c) 2005:                                                            *
 *      CERN, Switzerland                                                         *
 *      MPI-K Heidelberg, Germany                                                 *
 *      PAN, Krakow, Poland                                                       *
 *                                                                                *
 * Redistribution and use in source and binary forms, with or without             *
 * modification, are permitted according to the terms listed in LICENSE           *
 * (http://tmva.sourceforge.net/LICENSE)                                          *
 **********************************************************************************/
 
/*! \class TMVA::SVKernelMatrix
\ingroup TMVA
Kernel matrix for Support Vector Machine
*/
 
#include "TMVA/SVKernelMatrix.h"
 
#include "TMVA/MsgLogger.h"
#include "TMVA/SVEvent.h"
#include "TMVA/SVKernelFunction.h"
#include "TMVA/Types.h"
 
#include "RtypesCore.h"
 
#include <iostream>
#include <stdexcept>
 
////////////////////////////////////////////////////////////////////////////////
/// constructor
 
TMVA::SVKernelMatrix::SVKernelMatrix()
   : fSize(0),
     fKernelFunction(0),
     fSVKernelMatrix(0),
     fLogger( new MsgLogger("ResultsRegression", kINFO) )
{
}
 
////////////////////////////////////////////////////////////////////////////////
/// constructor
 
TMVA::SVKernelMatrix::SVKernelMatrix( std::vector<TMVA::SVEvent*>* inputVectors, SVKernelFunction* kernelFunction )
   : fSize(inputVectors->size()),
     fKernelFunction(kernelFunction),
     fLogger( new MsgLogger("SVKernelMatrix", kINFO) )
{
   fSVKernelMatrix = new Float_t*[fSize];
   try{
      for (UInt_t i = 0; i < fSize; i++) fSVKernelMatrix[i] = new Float_t[i+1];
   }catch(...){
      Log() << kFATAL << "Input data too large. Not enough memory to allocate memory for Support Vector Kernel Matrix. Please reduce the number of input events or use a different method."<<Endl;
   }
   // We compute the diagonal and one half of the off diagonal. When reading back we use
   // the symmetry of i,j to j,i to ensure the correct values are returned.
   for (UInt_t i = 0; i < fSize; i++) {
      for (UInt_t j = 0; j <=i; j++) {
         fSVKernelMatrix[i][j] = fKernelFunction->Evaluate((*inputVectors)[i], (*inputVectors)[j]);
      }
   }
}
 
////////////////////////////////////////////////////////////////////////////////
/// destructor
 
TMVA::SVKernelMatrix::~SVKernelMatrix()
{
   for (UInt_t i = fSize -1; i > 0; i--) {
      delete[] fSVKernelMatrix[i];
      fSVKernelMatrix[i] = 0;
   }
   delete[] fSVKernelMatrix;
   fSVKernelMatrix = 0;
}
 
////////////////////////////////////////////////////////////////////////////////
/// returns a row of the kernel matrix
 
Float_t* TMVA::SVKernelMatrix::GetLine( UInt_t line )
{
   Float_t* fLine = NULL;
   if (line >= fSize) {
      return NULL;
   }
   else {
      fLine = new Float_t[fSize];
      for( UInt_t i = 0; i <line; i++)
         fLine[i] = fSVKernelMatrix[line][i];
      for( UInt_t i = line; i < fSize; i++)
         fLine[i] = fSVKernelMatrix[i][line];
      return fLine;
   }
}
 
////////////////////////////////////////////////////////////////////////////////
/// returns an element of the kernel matrix
 
Float_t TMVA::SVKernelMatrix::GetElement(UInt_t i, UInt_t j)
{
   if (i > j) return fSVKernelMatrix[i][j];
   else       return fSVKernelMatrix[j][i]; // it's symmetric, ;)
}