[Rgc] [EXTERNAL] check of runs in low-PF epochs

[Sebastian Kuhn]

Hello again,

I did all of these “simplifications” of the original equations with the goal to ascertain whether my original assumption - namely, that it is better to calculate the PF first (a single, high-precision number for each epoch), and then the DF for each channel and kinematic bin within each epoch in terms of the packing fraction (depends only on epoch) and a factor containing only the auxiliary targets for each kinematic bin, divided by n_A. Turns out I was mistaken - there are cancellations in the original (straight-from-data) equation for the DF that don’t happen with my above approach, so even if the PF itself is infinitely well known, the uncertainty on the final DF in each bin would still be (slightly) larger with my previously preferred method than it is if one simply calculates DF directly from Eqs. 10 and 14. I have noticed during past meetings that some of you showed DF calculated both ways and the direct method showed smaller error bars than the PF-method. I didn’t understand that at the time, but I have now convinced myself that it is actually correct.

I still STRONGLY encourage everyone to calculate the PF - if nothing else, it can point to a problem with your specific channel and in any case serves as a nice check - and perhaps even the DF BOTH ways (with proper error propagation in both cases - for the PF-method, you can indeed assume 0 statistical uncertainty on PF but perhaps a 2% systematic uncertainty, so the only statistical uncertainties come from the auxiliary targets). Obviously, you want to combine all runs on each auxiliary target for each of the 5 sets - Summer 2022, Fall 2022 Solenoid+, Fall 2022 Solenoid-, Winter/Spring 2023 inbending, and Winter/Spring 2023 outbending (if possible). If one of the auxiliary targets is missing from one of these 5 sets, either use from the closest-in-kind set or by calculating ratios (e.g., CD2/CH2) from sets where they exist and then multiply (e.g., CH2 with that ratio).

Since I ALSO discovered a typo for ND3, maybe it would be best if you implemented the equations in the newest version (2.7) of the note, not only to make sure there are no mistakes, but also because they are simpler and more “straightforward” so there is less chance of a typo. If at all possible, it would be great to see some of these updated DF (and PF) calculations on Tuesday.

Sorry for the confusion and 180-degree turn on this, but it shows that sometimes intuition based on even a third of a century of experience can lead one to the wrong conclusion…

- Sebastian

On Feb 7, 2026, at 10:29, Sebastian Kuhn <kuhn at jlab.org> wrote:

Thank you for pointing this out. I updated the file again (v. 2.7) - hopefully now it is correct. (You may have to reload the Wiki page https://clasweb.jlab.org/wiki/index.php/Run_Group_C_Group_Meetings to get the updated link).

- Sebastian

On Feb 7, 2026, at 02:42, Hector Almanzor Chinchay Espino <HectorAlmanzor.ChinchayEspino at unh.edu> wrote:

Hello Sebastian,

Thanks for the updates, I saw a possible typo in this updated tech note in the equation of PF for ND3  (equation 15), I guess n_A should not be in the denominator in this case. Please, could you confirm this? thank you

Best,

Hector
________________________________
From: Sebastian Kuhn via Rgc_analysis <rgc_analysis at jlab.org<mailto:rgc_analysis at jlab.org>>
Sent: Friday, February 6, 2026 16:15
To: Silvia Niccolai <silvia at jlab.org<mailto:silvia at jlab.org>>
Cc: rgc_analysis at jlab.org<mailto:rgc_analysis at jlab.org> <rgc_analysis at jlab.org<mailto:rgc_analysis at jlab.org>>; rgc at jlab.org<mailto:rgc at jlab.org> <rgc at jlab.org<mailto:rgc at jlab.org>>
Subject: Re: [Rgc_analysis] [Rgc] check of runs in low-PF epochs


CAUTION: This email originated from outside of the University System. Do not click links or open attachments unless you recognize the sender and know the content is safe.


Dear RG-C analyzers,

I updated the tech note on dilution factor calculations based on data - see our Wiki page https://clasweb.jlab.org/wiki/index.php/Run_Group_C_Group_Meetings
I found a typo that affected the ND3 results, and I simplified all numerical expressions so they look more reasonable (and comparable to each other). Please check if you see any mistakes, and - if possible - re-calculate thePF and DF for your favorite channel and epoch using the PF method, preferably for both NH3 and ND3, but especially the latter.

Thanks, and happy weekend - Sebastian

On Jan 27, 2026, at 13:53, Silvia Niccolai via Rgc <rgc at jlab.org> wrote:

Dear all,
Following today's discussion, I contacted Chris to try and understand the ~10% differences that Hector observed in the PF for some ND3 epochs.
Chris suggested me to look at the raster distributions, so I went back to the Elog to retrieve the raster monitoring plots, and the results are in the attached slide. It seems like there is some vertical "sagging" of the material for the runs in the epochs having lower PF.
Best regards,
Silvia

<pf_differences_explanation.pdf>_______________________________________________
Rgc mailing list
Rgc at jlab.org<mailto:Rgc at jlab.org>
https://mailman.jlab.org/mailman/listinfo/rgc


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/rgc/attachments/20260207/3f37d178/attachment-0001.htm>