// @(#)root/tmva $Id$ // Author: Andreas Hoecker, Joerg Stelzer, Helge Voss /********************************************************************************** * Project: TMVA - a Root-integrated toolkit for multivariate data analysis * * Package: TMVA * * Class : ConvergenceTest * * Web : http://tmva.sourceforge.net * * * * Description: * * Contains all the data information * * * * Authors (alphabetical): * * Andreas Hoecker - CERN, Switzerland * * Peter Speckmayer - CERN, Switzerland * * Joerg Stelzer - CERN, Switzerland * * Helge Voss - MPI-K Heidelberg, Germany * * * * Copyright (c) 2006: * * CERN, Switzerland * * U. of Victoria, Canada * * 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://tmva.sourceforge.net/LICENSE) * **********************************************************************************/ #ifndef ROOT_TMVA_ConvergenceTest #define ROOT_TMVA_ConvergenceTest ////////////////////////////////////////////////////////////////////////// // // // ConvergenceTest // // // // check for convergence // // // ////////////////////////////////////////////////////////////////////////// #include #ifndef ROOT_Rtypes #include "Rtypes.h" #endif namespace TMVA { class ConvergenceTest { public: ConvergenceTest(); ~ConvergenceTest(); // setters void SetConvergenceParameters( Int_t steps, Double_t improvement ) { fSteps = steps; fImprovement = improvement; } void SetCurrentValue( Float_t value ) { fCurrentValue = value; } Float_t GetCurrentValue() { return fCurrentValue; } void ResetConvergenceCounter() { fCounter = -1; fMaxCounter = 0; } // getters Bool_t HasConverged( Bool_t withinConvergenceBand = kFALSE ); Float_t Progress(); // from 0 (just started) to 1 (finished) Float_t SpeedControl( UInt_t ofSteps ); protected: Float_t fCurrentValue; //! current value Float_t fImprovement; //! minimum improvement which counts as improvement Int_t fSteps; //! number of steps without improvement required for convergence private: Int_t fCounter; //! counts the number of steps without improvement Float_t fConvValue; //! the best "fitness" value Int_t fMaxCounter; //! maximum value for the counter so far // speed-control (gives back the learning speed = improvement-rate in the last N steps) // successList keeps track of the improvements to be able Float_t fBestResult; // Float_t fLastResult; // std::deque fSuccessList; // to calculate the improvement-speed }; } #endif