sPHENIX is a new detector at RHIC.

[Maxim Potekhin <potekhin AT bnl.gov>]

    Apologies, a wrong link was sent before. Here is the correct one:
    
    https://drive.google.com/file/d/1Y1aCPSLAs5C_N2U7DSqfGZ_ktCYM_tch/view?usp=sharing
    
    Maxim

On 3/16/2023 7:46 PM, Maxim Potekhin via sPHENIX-l wrote:

      Hello Dennis,
      
      Your comments are deeply appreciated! Thank you.
      
      This is the  link to the new updated version, corrected as per
      your suggestions:
      https://drive.google.com/file/d/1S_7QTwybKweuFdSxssSsuiWPLrZiyttD/view?usp=sharing
      
      Here are my itemized comments to your suggestions (which I took
      into consideration
      when correcting the paper, to a large extent) --

On 3/15/2023 9:49 PM, Perepelitsa, Dennis wrote:

        L41: “for event definition purposes” was a bit mysterious to me.
        I think the time is needed to reject signals not associated with
        a real beam-beam crossing, right? 
      
      
      Corrected. The language was indeed awkward, I've corrected it now.
      
      L62: I
        found myself curious about the specific type of interpolation.
        Was this a spline, polynomial, something else?
      
      It's actually linear -- it's the fastest, and given the factor 15
      oversampling we expected it to be adequate.
      And, I did a numerical comparison with a spline, and they match at
      more than 6 decimal places, so our expectation was
      correct. It's enough precision. I've added add a couple of words
      to elucidate that.

L91: would it be useful to give a citation or definition for the “modified Landau distribution” ?

      Done. There is a seminal paper where this functional form is
      derived for Landau, I've enclosed the citation.

L125-126: is there any potential issue with using a training sample with a distribution over energies that may not match the distribution of energies to which the ML method will be applied? For example, if the samples from the test beam are roughly similar for each energy, but in real data this will be a steeply falling distribution. Or is this mitigated by the template shape changing minimally with E?

      The dependency on the mix was there but not steep. I did some
      empirical tweaking of the mix and
      picked the one which produced the most precise inference. I have
      added a couple of words
      to clarify that situation.

L178 and L181: There is no Figure 4 - should this be Figure 3?

      This was a leftover from Tim and I tweaking the graphics, panels
      etc, it's corrected now.

Ref. 4: the ROOT authors have a requested citation - https://github.com/root-project/root/#cite .

      Done, thank you.

You might check if there is some specific way Refs. 5-9 like to be cited, rather than just as a website.

      These URLs are really the best sources. There is a strong
      technical angle to these, and this will be lost in
      a paper reference. Any conference presentation based on this won't
      be as good. If the referee requests that
      I'll change this.
      
      Thank you for the work you invested in reviewing this paper.
      Please let Tim and I know if you have further comments.
      We would like to close out by COB tomorrow, Friday, since the
      weekend is the deadline for the submission.
      
      Best regards,
      
      Maxim

On Mar 10, 2023, at 6:11 PM, Maxim Potekhin via sPHENIX-l <sphenix-l AT lists.bnl.gov> wrote:

Dear All,

Thanks to Dennis for pointing out my omission of not including the line numbers.
I created an additional copy with these included:
https://drive.google.com/file/d/1tCUCmqScIFHa7uaIY3YBhdIopK2UbrMK/view?usp=sharing
Please let me know if anything else needs fixing.
Best regards,
Maxim

On 3/10/2023 3:50 PM, Perepelitsa, Dennis wrote:

Hello Maxim,

Could you please circulate a version with line numbers? It may also be useful to copy the EMCal list.

Dennis

On Mar 10, 2023, at 3:42 PM, Maxim Potekhin via sPHENIX-l <sphenix-l AT lists.bnl.gov> wrote:

Dear Colleagues,

Tim Rinn and I have prepared a draft paper for the proceedings
of the ACAT 2022 Conference.

Title: "A Machine Learning Method for calorimeter signal processing in sPHENIX" The current author list: "M.Potekhin and T.Rinn for sPHENIX Collaboration"
We are asking for your comments for this draft publication, and will
greatly appreciate it if you send them by Thursday, March 16th, as the
submission deadline is the 19th.
The link to the PDF:
https://drive.google.com/file/d/1aZOO4a5bAQjha4bqBWAmnmcuyGIshQUv/view?usp=sharing
Thank you,

Maxim and Tim

_______________________________________________
sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov https://lists.bnl.gov/mailman/listinfo/sphenix-l

Dennis V. Perepelitsa
Associate Professor, Physics Department University of Colorado Boulder
on sabbatical at Brookhaven National Laboratory

_______________________________________________ sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov https://lists.bnl.gov/mailman/listinfo/sphenix-l

Dennis V. Perepelitsa
Associate Professor, Physics Department
University of Colorado Boulder
on sabbatical at Brookhaven National Laboratory

_______________________________________________
sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-l