sPHENIX tracking discussion

["W.A. Zajc" <zajc AT nevis.columbia.edu>]

Dear Joe, Tony, Jin, Hugo, Michael, Chris and Christof:

Thanks for this very interesting and informative write-up!

Below please find my comments. Most are stylistic rather than substantive, but there are some remarks at the end concerning the conclusions. 

Best regards,

Bill 

—————————————————

W.A. Zajc
I.I. Rabi Professor of Physics
Columbia University
New York, NY 10027

https://blogs.cuit.columbia.edu/waz1/
—————————————————

====================================================================================================

   8:  "will collect approximately 200 PB of data over three run periods utilizing a finite-sized computing center”

This reads strangely to me, since I am not aware of any infinite-sized computing centers.

(Please note: I am trying to be snarky here, just conveying why it caught my eye.)

Perhaps the sentence could be broken up along the lines of

“...will collect approximately 200 PB of data over three run periods. In order to make optimal use of available computing resources…”

 16: “5 second” —> “5 seconds”

 30: “Nuclear Physics” —> “Nuclear Science”

(a sensitive point for some people in the community)

 58: Suggest dropping “finite-sized”, as per above remarks.

 65: “detector hit occupancies  . . . O(100,000)” 

I am not sure what this means. This does not mean a change is required, just noting it’s not clear to me. I am used to expressing occupancies in percentage of available channels. Here you seem to be using it to express the number of discrete chunks of information per event (per crossing?). 

118: “approximately less than” sounds odd, suggest dropping “approximately” or replacing by “less than $\sim 1.2$%.”

Figure 2: Viewing the TPC in Mac Preview on my large monitor is not very informative - any possibility to adjust the shading?

164: If you have space, it would be interesting to describe how 3 degrees, rather than 1 or 10 degrees, was chosen. Presumably it was via some optimization based on pad plane segmentation, expected r-phi resolution, etc. 

185: First mention of Fun4All, reference?

Figure 3: “serves as a wrapper to that interfaces”       Drop “to”?

198: I know you won’t be able to revise this in time for the submission, but it seems somewhat odd after the previous discussion of > 1000 particles per event (not to mention pile-up) to present a result for 100 pions. I guess this is a proof of principle, but wonder how this degrades in the presence of higher multiplicities, kaon decays, hyperon decays, etc.

I do see this noted on line 235 regarding ongoing work, so please treat above as a comment rather than a suggestion for change. 

243: per MB or per central event?

Which raises the question - How does the time scale with the number of tracks? 

Figure 5: Not many in the readership will know what a “double sided crystal ball function” is. 

254: Again, the 10 seconds per event raises the question of scaling with the number tracks. 

347 “finite sized” again

CONCLUSION: The readership of CHEP proceedings is more inclined towards details about software than the mission goals of sPHENIX, which are a repeat of what was in the introduction. I suggest a couple of sentences noting a) Factor of 8 improvement in reconstruction time from initial implementation of ACTS, b) utility of the overall ACTS approach. The current conclusion emphasizes that incorporating TPC space-points is “nontrivial”, and I have every reason to believe this is the case, but in the interest of continuing good relations with the ACTS developers perhaps you could note that “while nontrivial, was made possible by the flexibility and uniform implementation of the ACTS tools” or anything else that makes it clear that the sPHENIX experience with moving to ACTS is a positive one. 

On May 11, 2021, at 2:27 PM, Osborn, Joe <osbornjd AT ornl.gov> wrote:

Hi all,

 

Attached is a draft of the tracking paper that is based on the proceedings I submitted to the CHEP conference. I implemented the reveiwer’s comments which were helpful and improved the paper; we will presumably get completely new reviewers once the paper is submitted. 

 

The target journal for submission is Computing and Software for Big Science. Please check your affiliation/name/etc to ensure that it is correct.

 

Please send any comments you might have by one week from today, that is, by May 18^th. This will give me a few days to implement the comments and then send the draft to the collaboration by May 23, which will then give us a few days before submission deadline to address any additional comments.

 

 

Joe Osborn

 

---------------------------

 

Joe Osborn, Ph.D.

Postdoctoral Research Associate

Oak Ridge National Laboratory

osbornjd AT ornl.gov

(859)-433-8738

 

<Implementation_of_ACTS_into_sPHENIX_Track_Reconstruction_SCBS.pdf>_______________________________________________
sPHENIX-tracking-l mailing list
sPHENIX-tracking-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-tracking-l