Global Particle ID Software in LHCb


This page contains technical information on the global Particle ID software in LHCb.


Sub-Detector Particle Identification in Brunel

Particle Identification starts with the sub detector specific reconstruction phases that run during normal processing in Brunel. These run algorithms specific to each sub-detector and store their final results in various summary objects in the T.E.S.


The main output of the RICH processing are RichPID data objects. These summarise the particle identification results for each track. The results are presented in the form a of delta log likelihood value for each mass hypothesis, that gives the overall change in the global event log likelihood when the track is question is changed from the pion mass hypothesis, the the hypothesis in question.

More details can be found on the RichSoftware pages.


The output from the Muon system PID algorithms are MuonPID data objects, which like the RICH summarise the PID information for each track.

More details can be found on the MuonSoftware pages.


* To Be Filled *

Particles and ProtoParticles

At the end of of the Brunel event processing, a final stage is run which forms ProtoParticle data objects. These objects are designed to represent a single reconstructed object (charged or neutral) in LHCB and to summarise all the quantities associated to that object. These ProtoParticles are stored in the final DSTs at the end of processing.

During processing in DaVinci, when the DSTs are read back in for physics analysis, one of the first steps that is need is to produce Particle candidates from the ProtoParticles. Each ProtoParticle can be made into zero, one or many ProtoParticles (with different PID types) depending on the PID information stored in the ProtoParticle and the PID selection cuts that are applied. This step is gnereally referred to as the 'ParticleMaker' step. For charged ProtoParticles, the CombinedParticleMaker is generally used, whilst for neutrals there are a selection of possible Particle Makers.

Gobal Particle ID

One task of the Global PID group is to study how best to merge all the sources of PID information (RICH, CALO, MUON etc.) to provide the best overall particle ID estimators.


Global PID was first introduced during the analysis of the DC04 data. The original study was performed by Roger Forty. The CombinedParticleMaker was first developed for the DC04 analyses. In this implementation, it was only possible to make selections based on cuts on the 'CombinedDLL' values. For futher information, see these emails from Roger.


During the transistion to the DC06 software framework, several changes where made to how the ProtoParticles are created and used :-

  • Firstly, they are now created (as originally intended) at the end of the processing in Brunel. In DC04 they are actually produced at the start of each event in DaVinci.
  • The amount of information stored in the ProtoParticle was extended.

ChrisRJones - 17 Nov 2006

This topic: LHCb > WebHome > LHCbComputing > GlobalParticleID > GlobalParticleIDSoftware
Topic revision: r4 - 2006-12-12 - unknown
This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback