search

umd-lhcb / lhcb-ntuples-gen

ntuples generation with DaVinci and in-house offline components
BSD 2-Clause "Simplified" License
1 stars 0 forks source link

Vertex smearing for MC #122

Open yipengsun opened 1 year ago

yipengsun commented 1 year ago

According to the ANA, final reweighting cannot correct for this effect, and these are included as separate templates.

The procedure is the following:

  1. (Greg did this, but can't find documentation anywhere) Do a 1D fit on the $m^2_{miss}$ to get a linear and quadratic correction factors for some angle randDeltaTheta (in Phoebe's notation)
  2. Re-compute rest frame variables based on the smeared kinematics. Available at: https://github.com/umd-lhcb/RDRDstRun1AnalysisPreservation/blob/7fc28713eb8ff359bd6a4a6f0f7d188118f78afd/proc/AddB.C#L4231-L4263 and https://github.com/umd-lhcb/RDRDstRun1AnalysisPreservation/blob/7fc28713eb8ff359bd6a4a6f0f7d188118f78afd/proc/AddB.C#L944-L960
  3. Add these smeared templates as variations for MC for every single one of them.
yipengsun commented 1 year ago

Sent an email to Greg asking about this.

yipengsun commented 1 year ago

Manuel asked the question that if we smear (change the angle between PV and B vertex) the B vertex, it would make sense that the track momenta forming the B vertex should also change.

This is considered unnecessary because:

  1. The vertex fit doesn't change daughter momenta much (In DaVinci, compare the regular vertex fitter vs. DecayTreeFitter, where the latter one re-adjust the track momenta of the vertex. The differences are small)
  2. Realistically, to change this, one has to do it at DaVinci level, which can be very hard to implement.

FYI @manuelfs

yipengsun commented 1 year ago

Related:

https://github.com/umd-lhcb/vertex-resolution/issues/1 https://github.com/umd-lhcb/vertex-resolution/issues/2

yipengsun commented 1 year ago

Vertex smearing disabled for now. Can be re-enabled by reverting 352c190f0c50309b3d7b37fb9023326169172e20.