eicrecon::TrackClusterMergeSplitter Class Reference

Track-Based Cluster Merger/Splitter. More...

#include <TrackClusterMergeSplitter.h>

Inheritance diagram for eicrecon::TrackClusterMergeSplitter:

Collaboration diagram for eicrecon::TrackClusterMergeSplitter:

Public Member Functions
	TrackClusterMergeSplitter (std::string_view name)
	! Algorithm constructor
void	process (const Input &, const Output &) const final
Public Member Functions inherited from eicrecon::WithPodConfig< TrackClusterMergeSplitterConfig >
TrackClusterMergeSplitterConfig &	getConfig ()
	Get a configuration to be changed.
TrackClusterMergeSplitterConfig &	applyConfig (TrackClusterMergeSplitterConfig cfg)
	Sets a configuration (config is properly copyible)

Additional Inherited Members
Public Types inherited from eicrecon::WithPodConfig< TrackClusterMergeSplitterConfig >
using	ConfigType = TrackClusterMergeSplitterConfig
Protected Attributes inherited from eicrecon::WithPodConfig< TrackClusterMergeSplitterConfig >
TrackClusterMergeSplitterConfig	m_cfg

Detailed Description

Track-Based Cluster Merger/Splitter.

An algorithm which takes a collection of clusters, matches track projections, and then decides to merge or split those clusters based on average E/p from simulations.

Heavily inspired by Eur. Phys. J. C (2017) 77:466

Constructor & Destructor Documentation

TrackClusterMergeSplitter()

eicrecon::TrackClusterMergeSplitter::TrackClusterMergeSplitter ( std::string_view  name )

inline

! Algorithm constructor

Member Function Documentation

process()

void eicrecon::TrackClusterMergeSplitter::process ( const Input &  , const Output &    ) const

final

Merges and splits clusters based on matched tracks according to the following algorithm:

1. Build map of clusters onto matched track projections.
2. For each cluster-track pair: a. Calculate significance of pair's E/p wrt provided average and RMS of E/p b. If significance is less than minSigCut, merge all clusters within drAdd.
3. Create a protocluster for each merged cluster
   • If multiple tracks point to same merged protocluster, create new protocluster for each projection with hits weighted relative to track.
4. Convert any unmatched clusters into protoclusters.

---

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

• src/algorithms/calorimetry/TrackClusterMergeSplitter.h
• src/algorithms/calorimetry/TrackClusterMergeSplitter.cc