Add TOPQ6 derivation format (slimmed TOPQ1 with custom track thinning)
This MR adds a new top derivation format, TOPQ6 to the derivation framework. It essentially follows everything done in TOPQ1, except:
- Custom (overall) track thinning defined via
InDetTrackParticles.pt > 0.1*GeV - Large-R and VR jet collections removed (and the related TCC and soft-drop jet collections)
Effectively, with respect to TOPQ1, this does not include the following containers:
AntiKt10TrackCaloClusterJets (DataVector<xAOD::Jet_v1>) [Jet]AntiKt10LCTopoJets (DataVector<xAOD::Jet_v1>) [Jet]AntiKt10TrackCaloClusterTrimmedPtFrac5SmallR20Jets (DataVector<xAOD::Jet_v1>) [Jet]AntiKt10TruthJets (DataVector<xAOD::Jet_v1>) [Jet]BTagging_AntiKtVR30Rmax4Rmin02Track (DataVector<xAOD::BTagging_v1>) [BTag]AntiKt10TruthTrimmedPtFrac5SmallR20Jets (DataVector<xAOD::Jet_v1>) [Jet]AntiKt10TruthSoftDropBeta100Zcut10Jets (DataVector<xAOD::Jet_v1>) [Jet]AntiKt10LCTopoCSSKSoftDropBeta100Zcut10Jets (DataVector<xAOD::Jet_v1>) [Jet]AntiKtVR30Rmax4Rmin02TrackJets (DataVector<xAOD::Jet_v1>) [Jet]AntiKt10LCTopoTrimmedPtFrac5SmallR20Jets (DataVector<xAOD::Jet_v1>) [Jet]
Due to the custom track thinning, the size of the InDetTrackParticles container is increased from about 12kb/events to about 37kb/event. The custom track thinning results in an overall increase of the derivation size by about 6%. Combined with the above containers removed, however, the TOPQ6 format is about 10% smaller than TOPQ1. All of these numbers are based on a small test I ran with a ttbar mc16a xAOD (410470) and 200 processed events. AnalysisTop runs fine over the produced TOPQ6 DAOD when removing the large-R selections in the default AnalysisTop validation-cuts.txt file. But it would probably be good to run some further tests that the format is complete and usable.
Closes TOPQDERIV-75