# Physics Performance Meeting

Slides: https://indico.belle2.org/event/2570/

## D* `kaonID` using the systematics framework (Sviat Bilokin)
* Based on and inspired by LHCb PIDCalib framework (https://arxiv.org/pdf/1803.00824.pdf)
* Produces ntuples with PID variables plus variables that drive PID performance (e.g. polar angle) plus the sWeights
* Workflow:
* Create offline skim
* Add ntuple with all variables
* Add PDF model to fit mass distribution
* Using KIT b2luigi framework to submit jobs to the grid
* Documentation is available here: https://kuhrios.universe-cluster.de/~bilokin/syscorrfw/

#### Questions

* Q Phillip [slide 8], do you mean the *systematics* of the method?

> A: no correction tables *or* systematic uncertainty
e.g. if you cut on `kaonID > 0.2` what is the systematic uncertainty on this cut.

> There should also be an uncertainty associated to the method (if you're using sweights etc).

* Q Torben: regarding the workflow, is the idea that every analyst who needs this correction contacts the liaison and he/she treuns the framework?

> A: it should be able to run automatically. Only need to contact the liaison if one needs to add somethng or there is a request for some improvement (new pdf model, new variable etc.)

* Q Bianca: is it based on the HLT_lambda skim? (thinking about Λ, the HLT_lambda skim currently has a tighter cut than the analysis skim... can this be a problem?)

> So far, the HLT skims will **not** be used for this. HLT_lambda is *currently* only used for Mirabelle. If this assumption changes we will need to revisit this.

> There are only offline skims available for analysts. These only run on all mdst or on HLT_hadron.

## Performance input to November B2GM and BPAC (Torben and liaisons).

* Plan for B2GM and BPAC is attached to agenda.

**Petar/Tracking**

* Confirmed the list is good.
* Plan to show updates for slow pions.
* R.e. tracking efficiency for non-IP tracks, let's see. Can push but no promises.
* High background monitoring/checks were done by Bianca, can discuss whether this is sufficient or if more needs to be done.
* Exact scope of the tracking part of the performance paper is being discussed.

**Savino/Neutrals**

* Not all of these topics are going to be in the neutrals paper.
* For π0/η decays there has been some delays. We have new personpower from Xinping.
* For Klong we are still looking for person power. Mostly simulation-based until now. Need to move to data.
* For E_extra, Racha is preparing a BDT. Should be ready for the B2GM. Not available for users yet.

**Phill/Lepton ID**

* PSD not yet used in BDT, will improve muon/pion. We want that included -- or at least tried out -- before we make it the default.
* paper planning (lead by Marco) ongoing
* problem with high momemtum electron/muon in muonID (reported by KLM())

**Hadron ID**
* Marco is porting FORTRAN code to C++
* Contrary to the hadron ID confluence page, Umberto does not work on TOP reconstruction.

**Mario/B counting**

* Can present updates for sure for the B2GM.
* Crucially need luminosity input for ststematics from off- and -on-resonance datasets. This is the biggest uncertainty for B-counting.
* Q: Does this mean you have communicated with Xing Yu?
* Yes... comunicated by email but still don't know timescale.

**Beam energy & boost**

* Work done by Renu and Jim Libby. Taken over by Hossin Ahmed from Canada.
* Potentially a separate paper, Jim is collecting interested people.

**Beam spot.**

* Historically coordinated by tracking group.
* [C: Petar] Nice updates lately from Radek on beam spot and boost. For sure this sould be updated at B2GM and BPAC.
* Since Radek looks also at boost Petar and Torben will follow up and coorinate with Hossin.

## AOB
* We will move the "asia friendly" meeting to 16:00 JST (instead of 18:00 JST) to make it also Australian-friendly.
* Europe switches to daylight saving time 25 Octover
* US/Canada 01 November