MLRange< ClassifierType, Ndims, CutType > Class Template Reference

Range used for the Machine Learning assisted TrackFinding approach. More...

#include <MLRange.h>

Public Member Functions
	MLRange (ClassifierType *classifier, CutType cut)
	Constructor.
bool	contains (std::array< double, Ndims > hits) const
	method used to decide if a hit combination passes the Machine Learning filter
void	persist (TTree *, const std::string &, const std::string &)
	For Storing this range.

Private Attributes
ClassifierType *	m_classifier
	Pointer to classifier to be used.
CutType	m_cut
	Cut value.

Detailed Description

template<typename ClassifierType, size_t Ndims = 9, typename CutType = double>
class Belle2::MLRange< ClassifierType, Ndims, CutType >

Range used for the Machine Learning assisted TrackFinding approach.

NOTE: This is where all the magic happens! For other Filters the SelectionVariable is responsible for calculating a value and then passes the calculated value to a Range, where a simple comparison occurs. In the case of ML Filters this is not easily possible if there should be different ML classifiers for different detector regions. Thus, the SelectionVariable simply passes down the hits and this Range is responsible for the calculation and the comparison.

Definition at line 29 of file MLRange.h.

Constructor & Destructor Documentation

MLRange()

template<typename ClassifierType, size_t Ndims = 9, typename CutType = double>

MLRange ( ClassifierType * classifier, CutType cut )

inline

Constructor.

Parameters

classifier	: classifier to be used
cut	: cut value

Definition at line 40 of file MLRange.h.

: m_classifier(classifier), m_cut(cut) { ; }

Member Function Documentation

contains()

template<typename ClassifierType, size_t Ndims = 9, typename CutType = double>

bool contains ( std::array< double, Ndims > hits ) const

inline

method used to decide if a hit combination passes the Machine Learning filter

Parameters

hits	is the three hit combination (for the moment, maybe this will get templated in the future) that gets passed down from the MLHandover selection variable.

Definition at line 48 of file MLRange.h.

{ return !(m_classifier->analyze(hits) < m_cut); }

persist()

template<typename ClassifierType, size_t Ndims = 9, typename CutType = double>

void persist ( TTree * , const std::string & , const std::string &  )

inline

For Storing this range.

TODO

Definition at line 55 of file MLRange.h.

    {
      // TODO
    }

Member Data Documentation

m_classifier

template<typename ClassifierType, size_t Ndims = 9, typename CutType = double>

ClassifierType* m_classifier

private

Pointer to classifier to be used.

Definition at line 31 of file MLRange.h.

m_cut

template<typename ClassifierType, size_t Ndims = 9, typename CutType = double>

CutType m_cut

private

Cut value.

Definition at line 33 of file MLRange.h.

---

The documentation for this class was generated from the following file:

• tracking/trackFindingVXD/filterTools/include/MLRange.h