sPHENIX EMCal discussion

["Perepelitsa, Dennis" <dvp AT bnl.gov>]

Hi all,

Just a thought on this:

> I was borrowing the term re-calibrator for the position-dependent energy 
> scale correction. I think it is a tool analyzer should individually choose 
> to use or not. For example, the Upsilon electron ID and photon-jet analysis 
> probably want to use it, as it sharpens the electron/photon response. The 
> EMCal+HCal cluster jet probably should NOT use it, as hadron shower in 
> EMCal will have different position response. 


I think that’s exactly right. For EMCal+HCal call tower jet reconstruction 
(the “default” in sPHENIX in some sense), I think one would want to apply 
such a calibration/correction at the jet-level after the fact of 
reconstruction. For a Particle Flow-style jet reconstruction, I believe one 
would calibrate/correct individual EMCal and HCal clusters before any jet 
finding, so in this case you would apply it cluster by cluster first.

Dennis 

> On May 31, 2017, at 9:00 AM, Huang, Jin <jhuang AT bnl.gov> wrote:
> 
> Hi Martin
> 
> Thank you for the comment!
> 
> The current production only produce and save the Geant4 information. 
> Towering, clustering and this position-dependent energy scale correction 
> are fast and are run by analyzers. Therefore, the cluster energy, corrected 
> or not are not going to the production DSTs anyway. 
> 
> I was borrowing the term re-calibrator for the position-dependent energy 
> scale correction. I think it is a tool analyzer should individually choose 
> to use or not. For example, the Upsilon electron ID and photon-jet analysis 
> probably want to use it, as it sharpens the electron/photon response. The 
> EMCal+HCal cluster jet probably should NOT use it, as hadron shower in 
> EMCal will have different position response. We probably want to see how 
> this can be applied to central Au+Au simulations, where background is high. 
> And the module should save the position-corrected cluster to an additional 
> EM cluster DST node, so analyzers could easily pick alternative ones and 
> compare. Shall we?
> 
> Cheers,
> 
> Jin
> 
> 
> 
> ______________________________
> 
> Jin HUANG
> 
> Associate Physicist
> Brookhaven National Laboratory
> Physics Department, Bldg 510 C
> Upton, NY 11973-5000
> 
> Office: 631-344-5898
> Cell:   757-604-9946
> ______________________________
> 
> 
>> -----Original Message-----
>> From: sPHENIX-EMCal-l [mailto:sphenix-emcal-l-bounces AT lists.bnl.gov] On
>> Behalf Of Martin Purschke
>> Sent: Wednesday, May 31, 2017 10:42 AM
>> To: sphenix-emcal-l AT lists.bnl.gov
>> Subject: Re: [Sphenix-emcal-l] Simulated Resolution of EMCal
>> 
>> Jin,
>> 
>> not disputing the recalibrator, but wouldn't such a has-to-be-applied
>> correction not go on the DSTs already, rather than armies of people
>> individually applying the correction after the fact? Just wondering...
>> the recalibrators were once designed to take care of things we discovered
>> after the fact, when a big production was done and a re-production just to
>> incorporate that was too expensive. It's not a lot of data, so we could 
>> just do
>> that.
>> 
>> My 5cts,
>> 
>>    Martin
>> 
>> 
>> On 05/31/17 09:53, Huang, Jin wrote:
>>> 
>>> Great Job! Joe!
>>> 
>>> 
>>> 
>>> It is also interesting that the energy resolution spread from
>>> different pseudo-rapidity bins become consistent after the recalibration.
>>> 
>>> 
>>> 
>>> Now it would be useful to introduce a module applying cluster energy
>>> recalibration for all analyzers of EM showers. It can be in the
>>> nightly build with the name of *coresoftware*
>>> <https://github.com/sPHENIX-Collaboration/coresoftware>/simulation
>>> <https://github.com/sPHENIX-
>> Collaboration/coresoftware/tree/master/sim
>>> ulation>/g4simulation
>>> <https://github.com/sPHENIX-
>> Collaboration/coresoftware/tree/master/simulation/g4simulation>/*g4cemc
>> */RawClusterRecalibrator.
>>> The calibration constant data file can be installed at  *calibrations*
>>> <https://github.com/sPHENIX-Collaboration/calibrations>/*CEMC**/**Cali
>>> bration**via pull requests too. An example of doing recalibration via
>>> calibration data files is from the last two test beams,
>>> **CaloCalibration*
>>> 
>> <https://www.phenix.bnl.gov/WWW/sPHENIX/doxygen/html/da/d36/classC
>> aloCalibration.html>*.
>>> Welcome to give a try.*
>>> 
>>> 
>>> 
>>> Cheers,
>>> 
>>> 
>>> 
>>> Jin
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> ______________________________
>>> 
>>> 
>>> 
>>> Jin HUANG
>>> 
>>> 
>>> 
>>> Associate Physicist
>>> 
>>> Brookhaven National Laboratory
>>> 
>>> Physics Department, Bldg 510 C
>>> 
>>> Upton, NY 11973-5000
>>> 
>>> 
>>> 
>>> Office: 631-344-5898
>>> 
>>> Cell:   757-604-9946
>>> 
>>> ______________________________
>>> 
>>> 
>>> 
>>> *From:*sPHENIX-EMCal-l [mailto:sphenix-emcal-l-bounces AT lists.bnl.gov]
>>> *On Behalf Of *Joe Osborn
>>> *Sent:* Wednesday, May 31, 2017 9:12 AM
>>> *To:* Haggerty, John <haggerty AT bnl.gov>
>>> *Cc:* sphenix-emcal-l AT lists.bnl.gov
>>> *Subject:* Re: [Sphenix-emcal-l] Simulated Resolution of EMCal
>>> 
>>> 
>>> 
>>> Dear EMCalers,
>>> 
>>> 
>>> 
>>> Following Jin's suggestion, I implemented a leading order position
>>> dependent recalibration on the energy response for photons, electrons,
>>> and positrons. The results are attached in this email.
>>> 
>>> 
>>> 
>>> Please note that the "before recalibration" plots are not the exact
>>> same as what I sent before as I was inadvertently only using 10% of
>>> the simulation that Chris produced (whoops). The results here are
>>> using the entire single particle production. The position dependent
>>> recalibration reduces the constant term by 1-2% for each eta bin.
>>> 
>>> 
>>> 
>>> Thanks,
>>> 
>>> 
>>> 
>>> Joe Osborn
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> ---------------------------------------
>>> 
>>> 
>>> 
>>> Joe Osborn
>>> 
>>> Ph.D Candidate
>>> 
>>> University of Michigan
>>> 
>>> jdosbo AT umich.edu <mailto:jdosbo AT umich.edu>
>>> 
>>> (859) 433-8738
>>> 
>>> 
>>> 
>>> On Sun, May 28, 2017 at 10:13 AM, Joe Osborn <jdosbo AT umich.edu
>>> <mailto:jdosbo AT umich.edu>> wrote:
>>> 
>>>    Hi John,
>>> 
>>> 
>>> 
>>>    Chris sent a couple of emails out to the software lists, but I
>>>    guess not the other lists. It might be too late as the BBQ might
>>>    already be fired up, but for anyone who wants to take a look the
>>>    G4 hits files are located in the following directory:
>>> 
>>> 
>>> 
>>>    /sphenix/data/data02/review_2017-08-02/
>>> 
>>> 
>>> 
>>>    Joe
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>>    ---------------------------------------
>>> 
>>> 
>>> 
>>>    Joe Osborn
>>> 
>>>    Ph.D Candidate
>>> 
>>>    University of Michigan
>>> 
>>>    jdosbo AT umich.edu <mailto:jdosbo AT umich.edu>
>>> 
>>>    (859) 433-8738 <tel:%28859%29%20433-8738>
>>> 
>>> 
>>> 
>>>    On Fri, May 26, 2017 at 4:04 PM, John Haggerty <haggerty AT bnl.gov
>>>    <mailto:haggerty AT bnl.gov>> wrote:
>>> 
>>>        Joe,
>>> 
>>>        Thanks to Joe, Jin, and Chris for these results.  There is
>>>        obviously some bouncy behavior that seems a little odd, but I
>>>        know that even in very simple cases, getting "the resolution"
>>>        at the few percent level is not as simple as it sounds.
>>> 
>>>        Just to be clear--this is with the right detector material
>>>        budget in front of the EMCAL, all events from (0,0,0), and
>>>        Sean's light collection map in every tower?
>>> 
>>>        Probably as you say nobody's going to have time between the
>>>        barbecue this weekend and the upcoming barbecues at the lab,
>>>        but you might ask Chris to post their location to the wiki and
>>>        the emcal and hcal lists.
>>> 
>>>        On 5/26/17 3:37 PM, Joe Osborn wrote:
>>> 
>>>            Dear EMCalers,
>>> 
>>>            As Jin presented in the simulations/EMCal meeting this
>>>            week, he has updated the EMCal simulations with the tower
>>>            tilting in azimuth and polar directions, amongst other
>>>            changes. With the simulations code tagged, Chris produced
>>>            single particle simulations that people can play with
>>>            before we start producing large batches of simulations.
>>> 
>>>            Attached are the resolution of the EMCal for photons,
>>>            electrons, and positrons in several different eta bins.
>>>            Just to be clear, the single particles were thrown for 9
>>>            different energies in the four different eta regions
>>>            surrounding the values quoted in my plots (i.e. eta=0
>>>            corresponds to truth particles being thrown in +/-0.05
>>>            units of pseudorapidity around 0, same for eta=0.3,0.6,
>>>            and 0.9).
>>> 
>>>            Any comments would be great, or perhaps everyone will be
>>>            too busy enjoying the long weekend in which case we can
>>>            also discuss them at the EMCal meeting next week.
>>> 
>>>            Thanks,
>>> 
>>>            Joe
>>> 
>>>            ---------------------------------------
>>> 
>>>            Joe Osborn
>>>            Ph.D Candidate
>>>            University of Michigan
>>>            jdosbo AT umich.edu <mailto:jdosbo AT umich.edu>
>>>            <mailto:jdosbo AT umich.edu <mailto:jdosbo AT umich.edu>>
>>>            (859) 433-8738 <tel:%28859%29%20433-8738>
>>> 
>>> 
>>>            _______________________________________________
>>>            sPHENIX-EMCal-l mailing list
>>>            sPHENIX-EMCal-l AT lists.bnl.gov
>>>            <mailto:sPHENIX-EMCal-l AT lists.bnl.gov>
>>>            https://lists.bnl.gov/mailman/listinfo/sphenix-emcal-l
>>> 
>>> 
>>> 
>>>        --
>>>        John Haggerty
>>>        email: haggerty AT bnl.gov <mailto:haggerty AT bnl.gov>
>>>        cell: 631 741 3358 <tel:631%20741%203358>
>>>        _______________________________________________
>>>        sPHENIX-EMCal-l mailing list
>>>        sPHENIX-EMCal-l AT lists.bnl.gov
>>>        <mailto:sPHENIX-EMCal-l AT lists.bnl.gov>
>>>        https://lists.bnl.gov/mailman/listinfo/sphenix-emcal-l
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> _______________________________________________
>>> sPHENIX-EMCal-l mailing list
>>> sPHENIX-EMCal-l AT lists.bnl.gov
>>> https://lists.bnl.gov/mailman/listinfo/sphenix-emcal-l
>> 
>> 
>> _______________________________________________
>> sPHENIX-EMCal-l mailing list
>> sPHENIX-EMCal-l AT lists.bnl.gov
>> https://lists.bnl.gov/mailman/listinfo/sphenix-emcal-l
> _______________________________________________
> sPHENIX-EMCal-l mailing list
> sPHENIX-EMCal-l AT lists.bnl.gov
> https://lists.bnl.gov/mailman/listinfo/sphenix-emcal-l

Dennis V. Perepelitsa
Assistant Professor, Physics Department
University of Colorado Boulder