MethodLD.h

Go to the documentation of this file.

// Author: Krzysztof Danielowski, Kamil Kraszewski, Maciej Kruk, Jan Therhaag 

/**********************************************************************************
 * Project: TMVA - a Root-integrated toolkit for multivariate data analysis       *
 * Package: TMVA                                                                  *
 * Class  : MethodLD                                                              *
 * Web    : http://tmva.sourceforge.net                                           *
 *                                                                                *
 * Description:                                                                   *
 *      Linear Discriminant (Simple Linear Regression)                            *
 *                                                                                *
 * Authors (alphabetical):                                                        *
 *      Krzysztof Danielowski <danielow@cern.ch> - IFJ PAN & AGH, Poland          *
 *      Kamil Kraszewski      <kalq@cern.ch>     - IFJ PAN & UJ, Poland           *
 *      Maciej Kruk           <mkruk@cern.ch>    - IFJ PAN & AGH, Poland          *
 *      Peter Speckmayer      <peter.speckmayer@cern.ch>  - CERN, Switzerland     *
 *      Jan Therhaag          <therhaag@physik.uni-bonn.de> - Uni Bonn, Germany   *
 *                                                                                *
 * Copyright (c) 2008-2011:                                                       *
 *      CERN, Switzerland                                                         * 
 *      PAN, Poland                                                               * 
 *      U. of Bonn, Germany                                                       *
 *                                                                                *
 * 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)                                          *
 *                                                                                *
 **********************************************************************************/

#ifndef ROOT_TMVA_MethodLD
#define ROOT_TMVA_MethodLD

//////////////////////////////////////////////////////////////////////////
//                                                                      //
// MethodLD                                                             //
//                                                                      //
// Linear Discriminant                                                  //
// Can compute multidimensional output for regression                   //
// (although it computes every dimension separately)                    //
//                                                                      //
//////////////////////////////////////////////////////////////////////////

#include <vector>

#ifndef ROOT_TMVA_MethodBase
#include "TMVA/MethodBase.h"
#endif
#ifndef ROOT_TMatrixDfwd
#include "TMatrixDfwd.h"
#endif

namespace TMVA {

   class MethodLD : public MethodBase {

   public:
   
      // constructor
      MethodLD( const TString& jobName, 
                const TString& methodTitle, 
                DataSetInfo& dsi,
                const TString& theOption = "LD",
                TDirectory* theTargetDir = 0 );
      
      // constructor
      MethodLD( DataSetInfo& dsi, 
                const TString& theWeightFile, 
                TDirectory* theTargetDir = 0 );

      // destructor
      virtual ~MethodLD( void );

      Bool_t HasAnalysisType( Types::EAnalysisType type, UInt_t numberClasses, UInt_t numberTargets );
    
      // training method
      void Train( void );

      // calculate the MVA value
      Double_t GetMvaValue( Double_t* err = 0, Double_t* errUpper = 0 );

      // calculate the Regression value
      virtual const std::vector<Float_t>& GetRegressionValues();

      using MethodBase::ReadWeightsFromStream;   

      void AddWeightsXMLTo      ( void* parent ) const;

      void ReadWeightsFromStream( std::istream & i );
      void ReadWeightsFromXML   ( void* wghtnode );

      const Ranking* CreateRanking();
      void DeclareOptions();
      void ProcessOptions();

   protected:

      void MakeClassSpecific( std::ostream&, const TString& ) const;
      void GetHelpMessage() const;

   private:

      Int_t fNRegOut; // size of the output
 
      TMatrixD *fSumMatx;              // Sum of coordinates product matrix 
      TMatrixD *fSumValMatx;           // Sum of values multiplied by coordinates
      TMatrixD *fCoeffMatx;            // Matrix of coefficients
      std::vector< std::vector<Double_t>* > *fLDCoeff; // LD coefficients

      // default initialisation called by all constructors
      void Init( void );

      // Initialization and allocation of matrices
      void InitMatrices( void );

      // Compute fSumMatx
      void GetSum( void );

      // Compute fSumValMatx
      void GetSumVal( void );

      // get LD coefficients
      void GetLDCoeff( void );
      
      // nice output
      void PrintCoefficients( void );

      ClassDef(MethodLD,0) //Linear discriminant analysis
         };
} // namespace TMVA

#endif // MethodLD_H