ROOT logo
ROOT » TMVA » TMVA::GeneticRange

class TMVA::GeneticRange


 Range definition for genetic algorithm

Function Members (Methods)

Data Members

private:
Double_tfFrom
TMVA::Interval*fIntervalholds the complete information of the interval
Int_tfNbins
TRandom*fRandomGeneratorthe randomGenerator for calculating the new values
Double_tfTothe constraints of the coefficient
Double_tfTotalLengththe distance between the lower and upper constraints

Class Charts

Inheritance Inherited Members Includes Libraries
Class Charts

Function documentation

GeneticRange(TRandom* rnd, TMVA::Interval* interval)
 defines the "f" (from) and "t" (to) of the coefficient
 and takes a randomgenerator

Double_t RandomDiscrete()
 creates a new random value for the coefficient; returns a discrete value

Double_t Random(Bool_t near = kFALSE, Double_t value = 0, Double_t spread = 0.1, Bool_t mirror = kFALSE)
 creates a new random value for the coefficient
 Parameters:
        Bool_t near     : takes a random value near the current value
        double value  : this is the current value
        double spread : the sigma of the gaussian which is taken to calculate the new value
        Bool_t mirror   : if the new value would be outside of the range, mirror = false
                        maps the value between the constraints by periodic boundary conditions.
                        With mirror = true, the value gets "reflected" on the boundaries.

Double_t ReMap(Double_t val)
 remapping the value to the allowed space

Double_t ReMapMirror(Double_t val)
 remapping the value to the allowed space by reflecting on the
 boundaries
~GeneticRange()
 destructor
GeneticRange(TRandom* rnd, TMVA::Interval* interval)
Double_t GetFrom()
{ return fFrom; }
Double_t GetTo()
{ return fTo; }
Double_t GetTotalLength()
{ return fTotalLength; }