Skip to content

RD developments towards April

Miguel Ramos Pernas requested to merge rd-devel into 2024-patches

Depends on lhcb-datapkg/ParamFiles!82 (merged).

Overall impact on bandwidth

Based on the following 2024-patches and this MR.

2024-patches this MR
FULL RD BW [GB/s] 0.16 0.16
TURBO RD BW [GB/s] 0.51 0.37
Sprucing for ALL lines (output to disk) [GB/s] 5.04 4.88
Hlt2 for ALL lines (output to tape) [GB/s] 11.4 11.33


Include modifications towards the deadline of April 12th, following !3053 (merged). Related to #710.

Prescales have been added to lines with a rate over 100 Hz from this test so they become smaller than 100 Hz. Line developers should take action and develop tighter selections in order to remove (or change) such prescales.

Modifications to be included:

  • Investigate possible variables with good separation power for hadronic tau lines, such as the relative distance between two vertices of the taus and/or any other vertex in the decay tree.
  • Study issues with lines doing too many combinations (like Hlt2RD_TauTo5Mu and Hlt2RD_TauTo7Mu), which significantly slow down the entire processing of HLT2.
  • Use 5-body combiners in Hlt2RD_TauTo5Mu and Hlt2RD_TauTo7Mu.
  • Append Full to full stream (radiative inclusive) lines, as in !2743 (merged))
  • Move B24l HLT2 lines to persistreco=true and remove corresponding sprucing lines (contingent on bandwidth numbers)
  • Include a new HLT2 line (Hlt2RD_TauToMuGamma_EE_MVA). Needs: lhcb-datapkg/ParamFiles!82 (merged)


If you want to contribute to this MR, please speak first to @mramospe and @matzeni and explain your goals. We should avoid having clashes across different files, so if your line is defined in a module where other lines live, please coordinate among the contributors to the file to avoid conflicts. Once you are given green light to do changes, please do so and include in the description of this MR a little summary indicating the file(s) that you have modified as well as the changes to the lines.

Please follow the same format as in !3053 (merged).



  • Added prescales to lines with rates over 100 Hz. The prescales have not been fine-tuned and set to multiples of 0.1 anticipating increases due to the higher multiplicity expected in real data. and (@jagoodin)

  • Hlt2RD_Displaced{4Mu|4E|MuMuEE}_Incl lines moved to persistreco=True
  • deleted (only contained sprucing equivalents to the lines in the point above) (@jagoodin, @fibishop )

  • Full appended to full-stream lines, as discussed below in !3157 (comment 7752236).
  • Add lines with DD converted photon and rename existing lines EE->EELL (@jagoodin) and rare_tau_decay_builders (@cacochat)

  • Included F.ISMUON requirement on any muons (whether PID cuts applied or not).
  • Moved 5mu and 7mu lines from 3+2(+2)-body combiners to 5(+2)-body combiners.
  • Include the new Hlt2RD_TauToMuGamma_EE_MVA line (@cacochat) and b_to_xtaul_builder (@tfulghes)

  • Add psedo-separation chi2 variable with cut log(tau_vtxsepchi2) > -1 (cc: @avenkate)
  • Looser pid cuts on pions
  • Tighter mass cuts on pion combinations (rhoMass from B->K*tautau analysis by Jacopo)
  • Tighter VTXCHI2 and BPVIPCHI2 cut on parent particle and (@mramospe)

  • Include a requirement on the vertex separation between intermediate particles in K_S^0\to X^0X^{(\prime)0} modes
  • Added some lines for rare \Omega^- decays
  • Tighten the requirements on photons and \pi^0 mesons to reduce multiple candidates
  • Add a requirement on the separation between the vertices of neutral intermediate particles in the exotic K_S^0 decays to avoid multiple candidates arising from swapping decay products between them
  • Reduce the prescales in normalization modes based on real data

b_to_xll_hlt2 (@rmwillia) b_to_xll_spruce (@rmwillia)

  • Remove electron PIDe from dictionaries, define explicitly when calling rdbuilder_thor.make_rd_detached_dielectron.
  • For BuToKpEE, BuToPipEE, B0ToKpPimEE, B0ToKpKmEE, B0ToPipPimEE (Same & Opposite sign) set PIDe>0 and tightened B-DIRA : > 0.9995 -> >0.9999 and vchi2pdof_max : > 9 to > 4.
  • For BuToKpPipPimEE, BuToKpKpKmEE (Same & Opposite sign) set PIDe>1.
  • For BtoKpResolvedPi0EE (Same & Opposite sign) PIDe : > 1 to > 2, PIDK : > 1 to > 2 [Rate was too high]
  • Corrected decay descriptor for B0ToKpKmEE, B0ToKpKmMuMu, B0ToPipPimEE, B0ToPipPimMuMu (Thanks to @jagoodin) to remove charge conjugation (i.e.,"[B0 -> J/psi(1S) rho(770)0]cc" -> "B0 -> J/psi(1S) rho(770)0").
  • There was a bug that, although in the PIDe was set in the selection dictionaries as > -4, because this was later called in and changed to > 1, it meant the PIDe cut that was actually applied was PIDe > 1.

b_to_xgamma_conv_exclusive_hlt2 and b_to_xgamma_conv_exclusive_spruce and b_to_xgamma_exclusive_hlt2 and b_to_xgamma_exclusive_spruce and rad_incl_builders (@fibishop)

  • Move lines with LL converted photon out of b_to_xgamma_exclusive modules into b_to_xgamma_conv_exclusive modules
  • Add lines with DD converted photon
  • Add possibility to cut on max PIDe in converted photon builders and and and and and and and and and and and (@lmadhanm)

  • Tightened vertex cuts on dihadron combinations as well as hadron+lepton combinations for LFV decays.
  • Reduced invariant Mass range for dihadron combinations as well as hadron+lepton combinations for LFV decays.
  • Tightened PID requirements for Kaon and Pion from Kstar and all electrons and muons

Global pardoning of lines with rates below 150Hz (@matzeni) The prescale on these lines was brought back to the status before c5ad661b .

Line Prescale used
Hlt2RD_BToPpPmEE 1
Hlt2RD_LbToPKTauTau_TauToE_SSe_OSpe 1
Hlt2RD_BdToKstTauTau_KstToKPi_TauToE_SSe_SSKe 1
Hlt2RD_BuToKTauTau_TauToE 0.15
Hlt2RD_LbToPKTauTau_TauToMu_FakeK 0.2
Hlt2RD_BdToKstTauTau_KstToKPi_TauToMu_SSKPi 1
Hlt2RD_BdToKstTauTau_KstToKPi_TauToMu_FakePi 1
Hlt2RD_LbToPKTauTau_TauToMu_TauToE_FakeMuon_OSpmu 0.04
Hlt2RD_LbToPKTauTau_TauToMu_TauToE_FakeKaon_OSpmu 0.4
Hlt2RD_BdToKstTauTau_KstToKPi_TauToMu_TauToE_SSKpi_OSKmu 0.5
Hlt2RD_BdToKstTauTau_KstToKPi_TauToMu_TauToE_FakeKaon_SSKmu 0.4
Hlt2RD_OmegamToXimPiPi_LLL 1
Hlt2RD_JpsiTo4E 1
Hlt2RD_UpsilonTo4E 1
Hlt2RD_UpsilonToMuE_SS 1
Edited by Miguel Ramos Pernas

Merge request reports