run_3_skim_tables.md

Proposed skim tables design

* = optional, only included as appropriate, ideally (need to confirm how to do this in O2 in a reasonable way). We should be able to turn all of the top level tables off through hyperloop to fit our needs

Event level

• Event table:
  • run number
  • some sort of trigger (eg. bool) or full trigger bit? - in favor of trigger bit word (full mask)
  • centrality* - yes
  • event plane?* - yes
  • vertex-Z position (pre-cut at 10cm)
  • information from FIT - T0, V0
  • any other quality cuts
    • flag for ambigous tracks - is this is also a per-track quantity?
    • number of HF, photon, J/psi candidates? (overlap with trigger information)

Tracks

• Track table for data:
  • pt (sign of pt of can be used to encode charge - yes)
  • eta
  • phi
• Track table for MC:
  • Same as above, plus:
  • ParticleID (when available)*
  • Label (ie. identifier between particle and true level)*
  • Optional encoded information field, where we can bit pack whatever else where want*
    • including track quality flag?

Clusters

• seperate table?
• Clusters table for data:
  • E
  • eta
  • phi
  • …?
• Clusters table for MC:
  • Same as approach as track table, minus PID?

Photons?

• note this goes to a separate table

HF

• separate table

• Photons:

  • Can we include candidates instead?