Skip to content
GitLab

InDetVKalVxInJetTool and JetTagTools, Retrained TrackClassificationTool BDT and ClassifiedTrackTagger implementation

Merged Katharina Voss requested to merge katharin/athena:ctt_tct_deco into 22.0

Modified the Track Classification Tool (TCT) implementation InDetVKalVxInJetTool/InDetTrkInJetType such to allow the usage of the retrained TCT BDT in five ptjet ranges, which shows better performance than the default TCT BDT. The retrained BDT used in InDetTrkInJetType (Track Classification Tool)

  • is deactivated by default (previously trained BDT ist used by default)
  • can be used instead of the default BDT by changing BTagging.TrkClassFiveBinMode to True in JetTagConfig/python/BTaggingConfigFlags.py

Further an AthAlgorithm and three member functions of InDetVKalVxInJetTool/InDetTrkInJetType are implemented to decorate xAOD::TrackParticle or xAOD::Jet objects with the TCT score. The newly implemented ClassifiedTrackTaggerTool was added under PhysicsAnalysis/JetTagging/JetTagTools, which is a b tagging algorithm particularly aimed at high ptjet and which uses the TCT information. Testing algorithms to check the decoration with TCT and CTT scores are added in InDetVKalVxInJetTool and in JetTagTools. This development was mainly part of my AQP, see AFT-562.

Edited by Katharina Voss