Skip to content

Performance Updates to RICH Reconstruction

Christopher Rob Jones requested to merge jonrob/Rec:RICH-likelihood-background-tuning-4D into master

Further tuning and bug fixes relating to the RICH 4D reconstruction for Upgrade2. Some improvements though also have a small benefit to the regular 3D (run3) reconstruction option.

  1. Some re-factoring of how the likelihood background algorithm works, now using PMT ECs as the grouping. Also adds some additional properties to allow more tuning for U2 luminosities and also when 4D is enabled. Due to these changes diffs are expected in tests that include the PID performance checkers.
  2. Improvements to the Long track emulation from MC. Now uses in addition to the RICH extended MC information from Gauss the tracking groups IdealStateProvider tool to create states at specific z positions using only tracking station MC hits. Current implementation is a bit run3 specific (i.e. it directly refers to VP,UT and FT hits) but in principle I think this could be improved and made to work with any future Upgrade2 tracking hits...
  3. Various minor additions and improvements to monitoring histograms and/or counters.
  4. A new track filter algorithm that applies an MC filtering to tracks, to select reconstructed tracks that have RICH extended MC information associated to them.

Requires LHCb!4337 (merged) Moore!2723 (merged) Lbcom!703 (merged)

Edited by Christopher Rob Jones

Merge request reports