IPruneTool.h

Go to the documentation of this file.

/**********************************************************************************
 * Project: TMVA - a Root-integrated toolkit for multivariate data analysis       *
 * Package: TMVA                                                                  *
 * Class  : TMVA::DecisionTree                                                    *
 * Web    : http://tmva.sourceforge.net                                           *
 *                                                                                *
 * Description:                                                                   *
 *      IPruneTool - a helper interface class to prune a decision tree            *
 *                                                                                *
 * Authors (alphabetical):                                                        *
 *      Doug Schouten <dschoute@sfu.ca> - Simon Fraser U., Canada                 *
 *                                                                                *
 * Copyright (c) 2005:                                                            *
 *      CERN, Switzerland                                                         *
 *      MPI-K Heidelberg, Germany                                                 *
 *                                                                                *
 * Redistribution and use in source and binary forms, with or without             *
 * modification, are permitted according to the terms listed in LICENSE           *
 * (http://mva.sourceforge.net/license.txt)                                       *
 **********************************************************************************/

#ifndef ROOT_TMVA_IPruneTool
#define ROOT_TMVA_IPruneTool

#include <iosfwd>
#include <vector>

#ifndef ROOT_TMVA_SeparationBase
#include "TMVA/SeparationBase.h"
#endif

#ifndef ROOT_TMVA_DecisionTree
#include "TMVA/DecisionTree.h"
#endif

namespace TMVA {

//    class MsgLogger;
  
   ////////////////////////////////////////////////////////////
   // Basic struct for saving relevant pruning information   //
   ////////////////////////////////////////////////////////////
   class PruningInfo {
    
   public:
    
      PruningInfo( ) : QualityIndex(0), PruneStrength(0), PruneSequence(0) {}
      PruningInfo( Double_t q, Double_t alpha, std::vector<DecisionTreeNode*> sequence );
      Double_t QualityIndex; //! quality measure for a pruned subtree T of T_max
      Double_t PruneStrength; //! the regularization parameter for pruning
      std::vector<DecisionTreeNode*> PruneSequence; //! the sequence of pruning locations in T_max that yields T
   };
  
   inline PruningInfo::PruningInfo( Double_t q, Double_t alpha, std::vector<DecisionTreeNode*> sequence )
      : QualityIndex(q), PruneStrength(alpha), PruneSequence(sequence) {}
  
   ////////////////////////////////////////////////////////////////////////////////////////////////////////////
   // IPruneTool - a helper interface class to prune a decision tree                                         //
   //                                                                                                        //
   // Any tool which implements the interface should provide two modes for tree pruning:                     //
   //   1. automatically find the "best" prune strength by minimizing the error rate on a test sample        //
   //      if SetAutomatic() is called, or if automatic = kTRUE argument is set in CalculatePruningInfo()    //
   //      In this case, the PruningInfo object returned contains the error rate of the optimally pruned     //
   //      tree, the optimal prune strength, and the sequence of nodes to prune to obtain the optimal        //
   //      pruned tree from the original DecisionTree                                                        //
   //                                                                                                        //
   //   2. a user-provided pruning strength parameter is used to prune the tree, in which case the returned  //
   //      PruningInfo object has QualityIndex = -1, PruneStrength = user prune strength, and PruneSequence  //
   //      is the list of nodes to prune                                                                     //
   ////////////////////////////////////////////////////////////////////////////////////////////////////////////
  
   class IPruneTool {
    
   public:

      typedef std::vector<const Event*> EventSample;
    
      IPruneTool( );
      virtual ~IPruneTool();
    
   public:
    
      // returns the PruningInfo object for a given tree and test sample
      virtual PruningInfo* CalculatePruningInfo( DecisionTree* dt, const EventSample* testEvents = NULL,
                                                 Bool_t isAutomatic = kFALSE ) = 0;
    
   public:
    
      // set the prune strength parameter to use in pruning
      inline void SetPruneStrength( Double_t alpha ) { fPruneStrength = alpha; }
      // return the prune strength the tool is using
      inline Double_t GetPruneStrength( ) const { return fPruneStrength; }
    
      // if the prune strength parameter is < 0, the tool will automatically find an optimal strength
      // set the tool to automatic mode
      inline void SetAutomatic( ) { fPruneStrength = -1.0; };
      inline Bool_t IsAutomatic( ) const { return fPruneStrength <= 0.0; }
    
   protected:
    
//       mutable MsgLogger* fLogger; //! output stream to save logging information
//       MsgLogger& Log() const { return *fLogger; }
      Double_t fPruneStrength; //! regularization parameter in pruning
    
    
      Double_t S, B;
   };
  
   inline IPruneTool::IPruneTool( ) :  
      fPruneStrength(0.0),
      S(0),
      B(0)
   {}
   inline IPruneTool::~IPruneTool( ) {}
  
}

#endif