sPHENIX discussion of software

["Huang, Jin" <jhuang AT bnl.gov>]

 

Calorimeter QA: Jin Huang

 

Slides 5:

Sarah: it appears the production data sample produced sharper distribution in sampling fraction than in Jin’s test production.

Jin: will follow up offline to zoom-in and quantify whether the difference is beyond statistics.  

ActionItem [Jin]: check into sampling fraction difference.

 

Part II: jet plots

 

Sarah: whether purity considered unique matching which is sensitive to the Clusterizer somehow split jet

Jin: not yet. Will apply.

 

Denis: suggest a few additional plots:

·       # of constituent in jet

·       Jet mass

·       Fraction of EMCal energy to total, and fraction of inner HCal, etc.

 

Denis : it will be useful to have an inefficiency plot

 

ActionItem [Jin]: update the pull request with above new suggestions.

 

Chris: how to capture a failure mode.

Jin: two layer of checking 1) QA module provide strict data structure check via Mike’s evaluator. If fail will abort production during tests. 2) summary plots give quick check against major problems. E.g. missing matching part of HCal leads to drop in jet reco efficiency quickly.

Chris: Suggest extract important numbers to provide quick global view

Jin: plan to do so when we have larger sample of QA data.

 

Ron: for electrons, only ~95% energy shows up in reconstructed cluster?

Jin: yes. Known problem in energy scale calibration at 10% level. To do better than that require much more careful calibration, e.g. sampling fraction VS eta.

 

 

HCal studies: Abhisek Sen

 

Part I: Observed up to 20% variation tilt to tile. Important for sPHENIX? Not much effect in energy response from variation of tile light collection efficiency.

 

Page3: it appears variation in tile light collection does not affect energy resolution much.

Jin: suggest fit energy dependent energy resolution in order to extract changes in constant term after applying tile variation.

 

ActionItem [Abhisek]: agree to do so.

 

John: it appears not much stuff affect HCal performance? Considering fewer plates per tower.

Jin: I would expect changes of sampling frequent (e.g. tile per tower) could change the statistical term more prominently.

 

 

Part II: preparing for implementing position dependent light collection efficiency

  

Page 5:

Jin: what is the size of variation

Ron: 40% to 60% from green and blue region

Jin: so same order of variation as the tile-to-tile light variation, so we could expect similar level of effect on the energy resolution.

 

Abhisek: request saving local coordinates to HCal G4hits.

Jin: suggest add as tag to the property list to G4hit

ActionItem [Abhisek and Chris] : follow up with Chris in implementing saving the local coordinate

 

Jin: if stretch four corners to match, would there be big difference between light collection maps of the central tile and the outer tile.

Ron/John: no outer tile produced yet. This is a good guess. Will measure if we receive them.

 

 

Data formats for test beam: John Haggerty

 

Page 9:

John: for the multi-sample analysis with our 30ns shaping time, peak fit and integrated energy are linearly correlated for cosmic data

Jin: from simulation, expect observing difference of these two at timing scale of 100ns level for developed low e/h-ration hadron shower. Use peak fit as energy could lead to higher e/h ratio.

John: good to check in data.

 

John: few cautious, e.g. page8:  

·       Use time stamp (ns) or sample unit?

·       Signal could be positive or negatively polarized depending on cabling

·       Need channel map

·       Signal shape fit gives better result although could take long time to be available

·       T0 jitter around since our DAQ clock is asynchronous with beam bunch at FermiLab

·       Calibration map

 

Jin: what is the ADC / pixel conversion?

John: 10 pixel per ADC with low gain configuration. 1/16 lower for high gain configuration.  

 

Chris: concern on RC DAQ use a different format of PRDF. Should use one format for sPHENIX software.

ActionItem [Martin and help]: clarify sPHENIX PRDF format. Implement event lib to support read of test beam data. 

 

 

Fast calorimeter simulation for tracking: Kurt Hill

 

Kurt: testing tracker configurations which required fast calorimeter simulation to assist fake rejection without too much CPU time in calorimetry G4.

 

Page 4:

Jin: suggest use segmentation of 0.1x0.1 in deta x dphi, which mimic the segmentation of HCal, that is mainly used to measure the hadron shower energy.

Jin: suggest remove the digitizer layer too.

Chris: suggest build special layer record energy flux at each radius of the calorimeters.

 

Good progress. Expect update next week.

 

______________________________

 

Jin HUANG

 

Brookhaven National Laboratory

Physics Department, Bldg 510 C

Upton, NY 11973-5000

 

Office: 631-344-5898

Cell:   757-604-9946

______________________________

 

 

> -----Original Message-----

> From: Huang, Jin

> Sent: Monday, February 8, 2016 1:16 PM

> To: 'sphenix-l AT lists.bnl.gov' <sphenix-l AT lists.bnl.gov>; 'sphenix-software-

> l AT lists.bnl.gov' <sphenix-software-l AT lists.bnl.gov>

> Subject: [one-time room 3-191] sPHENIX simulation and software meeting,

> Tue Feb 9th, 1PM EST @ 3-191

>

> Dear All,

>

> This is a reminder that we will have a brief sPHENIX simulation and software

> meeting, Tue Feb 9th, 1PM EST @ 3-191. We will try to finish before the start

> of RBRC workshop at 2PM.

>

> Currently on the agenda: following software and production update, I will try

> to wrap up the last part of calorimeter QA. Also we may have more

> discussion on the prototype beam test software preparation.

>

> The agenda page is at

> https://indico.bnl.gov/conferenceDisplay.py?confId=1667    , for which the

> common modification password is "1008"

> •            For residents at BNL, we will meet at the 3-191 conference room this

> time

> •            To join the Meeting via phone bridge:

> https://bluejeans.com/619960505

> •            To join via Phone:

> 1) Dial:

>         +1.408.740.7256

>         +1.888.240.2560(US Toll Free)

>         +1.408.317.9253(Alternate number)

>         (see all numbers - http://bluejeans.com/numbers)

> 2) Enter Conference ID: 619960505

>

> Cheers,

>

> Jin

>

> ______________________________

>

> Jin HUANG

>

> Brookhaven National Laboratory

> Physics Department, Bldg 510 C

> Upton, NY 11973-5000

>

> Office: 631-344-5898

> Cell:   757-604-9946

> ______________________________