Draft: [QEE] ew_projects_Run3 : MC Production Options + Tag&Probe PV Constraints
Supersedes many (all) files from these 3 APs: !1448 (merged), !1633 (merged), !1653 (merged)
Changes
Starting Point : v1r1448p2 (!1653 (merged))
!1549 (1de3d7b5) + !1549 (b9488eff) + !1549 (616573ae)
Collect relevant commits from !1448 (merged), v1r1448p1 (!1633 (merged)) and v1r1448p2 (!1653 (merged))
Aesthetic changes / Preparation changes
!1549 (d726ec66) + !1549 (e02c2013)
These are minor clean-ups that were useful for the Tag&Probe recombination change but I separated them to make the diff easier to review
make Tag&Probe candidates
Using the rawbanks persisted in the turcal stream that our ZTrkEff lines are in (PersistReco_RawBanks), we can make the tag and probe pairs and combine them at will. -> This is the easiest way to constrain the probe PVs to be those of the tag PVs.
remove keep_running from 24c1 + 24c2
As these productions are finished, we want to process all the data in those bookkeeping directories (and hadd the files only once all of them are processed, rather than 'taking shots in the dark' with keep_running).
Use Moore/v55r11 for 24c2 MC processing
Althought v55r10p1 and v55r11 were both used for 24c2, v55r11 was used for more of it (at least, for all of the data with the good alignment..., so if we have to choose we should chose v55r11).
New Variables
Simple boolean to flip on the nCandidate/totCandidate variables, I noticed we use this in the ew-analyses Run2 jobs so thought easy enough to include and expect little change to storage size.
ToDo
-
Check the TrkEff Mass between standard and PVConstrained on a suitable number of MC - Using a single LFN doesn't provide enough suitable events, while running over several breaks the memory limit -> will have to test by hand.
WishList
-
Constrain PVs of Tag&Probe samples -
ZVeto for Z-like candidates (not just W-like). - for correct ZVeto weights
-
Access variables like mu_bare_pt