sPHENIX MAPS tracker discussion

[Hideki Okawa <Hideki.Okawa AT cern.ch>]

Joint Heavy Flavor & Quarkonium Topical Group Meeting (2021/10/27)
https://indico.bnl.gov/event/13633/

Introduction

Brief updates from Cameron on the production plans

Electron reco in the EMCal - Sasha Bazilevsky

p.3

Tony: Electrons radiate mostly in INTT. The tail does not matter

  much, but the core of the peak is important for us.

p.5

Tony: The centrality dependence is ugly. How can we calibrate this

  away?

Sasha: Using Ecore will reduce the centrality dependence at least.

p.7

Tony: This is not after the Ecore cut, right?

Sasha: Right. The Ecore version is in the backup (p.12). The impact

  of the chi^2 cut will be reduced with Ecore.

p.8

Marzia: We can recalibrate the bias for the electrons in the bottom

  plot, right?

Sasha: Yes, it can be directly corrected, but the resolution stays

  worse.

Tony: Why the resolution becomes worse?

Sasha: The electrons hit the calorimeter with large angles. That

  introduces fluctuations in the longitudinal direction.

Marzia: We can recalibrate as we did in PHENIX, and the efficiency

  would stay the same. The resolution effect should not affect us.

Sasha: That is true.

Marzia: One concern is the eta dependence. Around eta~0, it would

  be PHENIX-like, but in larger eta, things would be different.

Sasha: I will prepare something later.

Marzia: The electron efficiency loss should be checked after the

  track-cluster matching.

Tony: Right. The impact is not clear to me.

Hideki: What is the plan for the Gaussian Sum Filter (GSF), btw?

Tony: Not the top of the list, but yes, of course we should consider it.

  GSF exists in ACTS, but we are not sure how well it is implemented.

  Once implemented, E/p may behave worse with GSF, but we will have

  to see.

Updates on vertexing & D0 pileup performance - Sanghoon Lim

Tony: How is the fitting algorithm done for RAVE? In our code, we only

  add tracks to the vertex list included in the fit. The rest of the tracks

  will be associated to the closest vertex later.

S.H.: For RAVE, all tracks will be considered in the fit and a weight will

  be assigned to each track.

Tony: I hope to update the code to be able to handle multiple time bins.

  It should be done in 1-2 months from now.

S.H.: Is there an easy way to use the SimpleVertexFinder in the analysis

  module? I want to try it for the b-tag.

Tony: The SimpleVertexFinder is the default, so it would be straightforward

  to use it. If we run RAVE, we can also store a vertex map for it.

  However, RAVE is too slow to consider for the production.