The Trimmed Kalman vertex reconstructor

Complete:

Overview

The Trimmed Kalman vertex reconstructor iteratively uses the Trimmed Kalman vertex fitter to find and fit interaction vertices. After each fitted vertex, the remaining tracks are re-used to search for further vertices.

The fitter is implemented in the package RecoVertex/TrimmedKalmanVertexFinder

Usage

A default An easy way to use the TKVF is through the KalmanTrimmedVertexFinder. It uses a conventional Kalman filter to fit the vertices, without the user having to specify anything. It is then used like any other VertexReconstructor.

KalmanTrimmedVertexFinder finder;
vector<TransientVertex> vertices = finder.vertices ( vector<TransientTrack> tracks );

Parameters

To change the parameters, the easiest is to use the setParameters(ParameterSet) method. Alternatively, individual set methods can be used. The PSet needed is the following:

GSFParameters = cms.PSet(
   ptCut = cms.double(0.0),
   vtxFitProbCut = cms.double(0.01),
   trackCompatibilityToPVcut = cms.double(0.05),
   trackCompatibilityToSVcut = cms.double(0.01),
   maxNbOfVertices = cms.int32(0)
)

The parameters are the following:

Parameter Name	Description	Default	method
maxNbOfVertices	the maximum number of vertices searched for. 0 means no limitations	0	setMaxNbOfVertices(int)
ptCut	the minimum pT (in GeV) of the tracks used to make vertices.	1.5	setPtCut(float)
trackCompatibilityToPVcut	the probability below which a track is considered incompatible with the 1st vertex candidate formed.	0.05	setTrackCompatibilityCut(float)
trackCompatibilityToSVcut	the probability below which a track is considered incompatible with the next vertex candidates formed.	0.01	setTrackCompatibilityToSV(float)
vertexFitProbabilityCut	the probability below which a vertex is rejected	0.01	setVertexFitProbabilityCut(float)

-- Main.speer - 05 Jul 2007