[PWGHF] Add centrality information to MC collisions #13940
Conversation
1, Added centrality for MC collisions and updated relevant event split and event loose processing functions.
|
O2 linter results: ❌ 0 errors, |
github-actions
bot
changed the title
Tao-Fang
requested review from
NicoleBastid,
alibuild,
apalasciano,
deepathoms,
fcatalan92,
fcolamar,
fgrosa,
gluparel,
hahassan7,
jpxrk,
mfaggin,
mmazzilli,
stefanopolitano,
vkucera and
zhangbiao-phy
as code owners
| #include "PWGHF/Core/HfHelper.h" | ||
| #include "PWGHF/DataModel/CandidateReconstructionTables.h" | ||
| #include "PWGHF/DataModel/CandidateSelectionTables.h" | ||
| #include "PWGLF/DataModel/mcCentrality.h" |
There was a problem hiding this comment.
Hi @Tao-Fang ! Thanks for the development. This is fine for me, I just have a doubt about this line: do you really need to include the centrality from LF? Could you not use the centrality table from HF?
There was a problem hiding this comment.
Hi @stefanopolitano, thanks for your comment, I think the structure with mccentrality seems fine, which has same cite on other HF codes like candidateCreator3Prong.cxx, candidateCreatorCascade.cxx, and taskOmegac0ToOmegaPi.cxx
There was a problem hiding this comment.
Hi @stefanopolitano, the centrality from the LF task assigns a value to all generated events (also those not reconstructed) according to the generated particles in the FT0 acceptance, and it has been calibrated to match the data. This information is needed, e.g., to compute the event loss as a function of multiplicity, where you need the multiplicity interval also for events not reconstructed.
What is done in HF is to assign to the generated collision the centrality of the associated reconstructed collision with the larger number of contributors (to address split vertices). BTW, I'm not sure how much we can trust the centrality intervals obtained in this way, given also that we are using enriched HF MCs.
There was a problem hiding this comment.
Hi @fcatalan92! Thanks for the clarification (and sorry for the delay). Good for me!
3 participants
1, Added centrality for MC collisions and updated relevant event split and event loose processing functions.