EIC code displayed by LXR |
|
|
|
|||
File indexing completed on 2025-01-30 10:22:53
0001 // @(#)root/tmva $Id$ 0002 // Author: Andreas Hoecker, Xavier Prudent, Joerg Stelzer, Helge Voss, Kai Voss 0003 0004 /********************************************************************************** 0005 * Project: TMVA - a Root-integrated toolkit for multivariate data analysis * 0006 * Package: TMVA * 0007 * Class : MethodFisher * 0008 * * 0009 * * 0010 * Description: * 0011 * Analysis of Fisher discriminant (Fisher or Mahalanobis approach) * 0012 * * 0013 * Original author of this Fisher-Discriminant implementation: * 0014 * Andre Gaidot, CEA-France; * 0015 * (Translation from FORTRAN) * 0016 * * 0017 * Authors (alphabetical): * 0018 * Andreas Hoecker <Andreas.Hocker@cern.ch> - CERN, Switzerland * 0019 * Xavier Prudent <prudent@lapp.in2p3.fr> - LAPP, France * 0020 * Helge Voss <Helge.Voss@cern.ch> - MPI-K Heidelberg, Germany * 0021 * Kai Voss <Kai.Voss@cern.ch> - U. of Victoria, Canada * 0022 * * 0023 * Copyright (c) 2005: * 0024 * CERN, Switzerland * 0025 * U. of Victoria, Canada * 0026 * MPI-K Heidelberg, Germany * 0027 * LAPP, Annecy, France * 0028 * * 0029 * Redistribution and use in source and binary forms, with or without * 0030 * modification, are permitted according to the terms listed in LICENSE * 0031 * (see tmva/doc/LICENSE) * 0032 **********************************************************************************/ 0033 0034 #ifndef ROOT_TMVA_MethodFisher 0035 #define ROOT_TMVA_MethodFisher 0036 0037 ////////////////////////////////////////////////////////////////////////// 0038 // // 0039 // MethodFisher // 0040 // // 0041 // Analysis of Fisher discriminant (Fisher or Mahalanobis approach) // 0042 // // 0043 ////////////////////////////////////////////////////////////////////////// 0044 0045 #include <vector> 0046 0047 #include "TMVA/MethodBase.h" 0048 #include "TMatrixDfwd.h" 0049 0050 class TH1D; 0051 0052 namespace TMVA { 0053 0054 class MethodFisher : public MethodBase { 0055 0056 public: 0057 0058 MethodFisher( const TString& jobName, 0059 const TString& methodTitle, 0060 DataSetInfo& dsi, 0061 const TString& theOption = "Fisher"); 0062 0063 MethodFisher( DataSetInfo& dsi, 0064 const TString& theWeightFile); 0065 0066 virtual ~MethodFisher( void ); 0067 0068 virtual Bool_t HasAnalysisType( Types::EAnalysisType type, UInt_t numberClasses, UInt_t numberTargets ); 0069 0070 0071 // training method 0072 void Train( void ); 0073 0074 using MethodBase::ReadWeightsFromStream; 0075 0076 // write weights to stream 0077 void AddWeightsXMLTo ( void* parent ) const; 0078 0079 // read weights from stream 0080 void ReadWeightsFromStream( std::istream & i ); 0081 void ReadWeightsFromXML ( void* wghtnode ); 0082 0083 // calculate the MVA value 0084 Double_t GetMvaValue( Double_t* err = nullptr, Double_t* errUpper = nullptr ); 0085 0086 enum EFisherMethod { kFisher, kMahalanobis }; 0087 EFisherMethod GetFisherMethod( void ) { return fFisherMethod; } 0088 0089 // ranking of input variables 0090 const Ranking* CreateRanking(); 0091 0092 // nice output 0093 void PrintCoefficients( void ); 0094 0095 0096 protected: 0097 0098 // make ROOT-independent C++ class for classifier response (classifier-specific implementation) 0099 void MakeClassSpecific( std::ostream&, const TString& ) const; 0100 0101 // get help message text 0102 void GetHelpMessage() const; 0103 0104 private: 0105 0106 // the option handling methods 0107 void DeclareOptions(); 0108 void ProcessOptions(); 0109 0110 // Initialization and allocation of matrices 0111 void InitMatrices( void ); 0112 0113 // get mean value of variables 0114 void GetMean( void ); 0115 0116 // get matrix of covariance within class 0117 void GetCov_WithinClass( void ); 0118 0119 // get matrix of covariance between class 0120 void GetCov_BetweenClass( void ); 0121 0122 // and the full covariance matrix 0123 void GetCov_Full( void ); 0124 0125 // get discriminating power 0126 void GetDiscrimPower( void ); 0127 0128 // get Fisher coefficients 0129 void GetFisherCoeff( void ); 0130 0131 // matrix of variables means: S, B, S+B vs. variables 0132 TMatrixD *fMeanMatx; 0133 0134 // method to be used 0135 TString fTheMethod; ///< Fisher or Mahalanobis 0136 EFisherMethod fFisherMethod; ///< Fisher or Mahalanobis 0137 0138 // covariance matrices 0139 TMatrixD *fBetw; ///< between-class matrix 0140 TMatrixD *fWith; ///< within-class matrix 0141 TMatrixD *fCov; ///< full covariance matrix 0142 0143 // number of events (sumOfWeights) 0144 Double_t fSumOfWeightsS; ///< sum-of-weights for signal training events 0145 Double_t fSumOfWeightsB; ///< sum-of-weights for background training events 0146 0147 std::vector<Double_t>* fDiscrimPow; ///< discriminating power 0148 std::vector<Double_t>* fFisherCoeff; ///< Fisher coefficients 0149 Double_t fF0; ///< offset 0150 0151 // default initialisation called by all constructors 0152 void Init( void ); 0153 0154 ClassDef(MethodFisher,0); // Analysis of Fisher discriminant (Fisher or Mahalanobis approach) 0155 }; 0156 0157 } // namespace TMVA 0158 0159 #endif // MethodFisher_H
| [ Source navigation ] | [ Diff markup ] | [ Identifier search ] | [ general search ] |
|
This page was automatically generated by the 2.3.7 LXR engine. The LXR team |
|