Hexaquarks optimisation

tighten selections in hexaquarks.py module to reduce rate/BW.

• Ps/Hs/Hss lines:

  • tighten cut on LTIME from 0.25ps to >2.0ps
  • remove prescale from HsToPPPi

• tighten PsToPpPimPim line (20-30% of BW) :

  • probnnp > 0.7
  • pt_p > 500 MeV
  • pt_sum > 1.3 GeV

• split all lines into RS and WS, put additional prescale of 0.5 of those WS lines with more than 1 decay mode

• replace PID cuts with PROBNN:

  • protons: PID_P > 5 -> PROBNN_P > 0.25, PID_P - PID_K > 0 -> PROBNN_K < 0.9
  • kaons: PID_K > 0 -> PROBNN_K > 0.1
  • pions: PID_K < 5 -> PROBNN_PI > 0.1
  • for detached pp modes: for at least one PID_P-PID_K > 5.0 -> PROBNN_P > 0.8 & PROBNN_K < 0.8

• tighten selection for Ks and Lambda on top of those in standart_particles:

  • PT(pi)>100 MeV, PT(p)>400 MeV
  • PT(Ks)>400(500) MeV, PT(Lambda)>800(1000) MeV for LL(DD)
  • VCHI2DOF < 10
  • given tighter cuts remove additional 0.5 prescale from modes with Lambda

• Tighten selection on all exclusive Hq lines:

  • PT(p)>200 -> >350 MeV
  • H_BPVIPCHI2 < 25 -> <9
  • sum_PT > 1 GeV -> > X, where X ~ 0.65 * expected mass of Hq

• detached pp and LamP lines:

  • replace full PersistReco with saving only Long(Downstream)Protos and Long(Downstream)Tracks
  • relax IPCHI2 < 100 cut given that's its only partially reconstructed thing

The Hlt2 output BW is reduced from 43 MB/s to 8.2 MB/s.