Thanks for posting these plots to confluence and letting everyone know. I have a lot of questions and comments regarding
what you are showing. What were the various selections used in these plots? Are there trigger requirements ever? It is hard
to draw many conclusions from looking at these without knowing for sure what was done. Could you share the code you
use to make the plots and explain how you configure and run this code to produce these plots so our collaborators can
reproduce these results and check the code/configuration directly to better understand what exactly they are looking at?
The momentum vs tan lambda plots are interesting to me. The shapes hint to me that we are possibly reconstructing
low tan lambda FEEs with a lower momentum than we should in a systematic way. I think it would be interesting to study more
directly the low tan lambda region. The kinda "sharp" drop in rate at low tan lambda of tracks at about the beam energy is
pretty well lined up with where the low momentum peak is turning on in tan lambda. It is hard to draw conclusions without
clearly following what the selection is and how these plots are produced. The way the code is implemented is important
to understand. There is a lot going on here so I would like to focus really just on 14168 first because this is a run I personally
already look at a lot.
I would be interested in seeing one particular region of what you are showing in a more isolated way in comparison to what
you already have in your slides.
The lower left-hand part of the plot is the region I was referring to earlier in this email. Could you make some cuts in momentum
and tan lambda to isolate that particular region as low p and low tan lambda? Then I would like to see a region with the same
cut on tan lambda (say around 0.025) but with an inverted p cut (so the higher momentum blob at low tan lambda). Then I would
like to define a third region which is high momentum and high tan lambda (inverted tan lambda cut). I would be really interested in
seeing all the same cluster position differences vs cluster x plots you show (including the overlap region, also interested in EXACTLY
how you define this in the code, the presentation does not appear to define it) but for all three of these regions and compare these
to the unchopped up selection you are already showing (that I don't understand). I think this could help us isolate where we are studying
so we can come to solid conclusions as to what exactly could be causing the reconstruction issues you are showing. I think it could also be
helpful to see these difference plots wrt the track x position and not just the cluster x position. I would also be interested in looking at an
additional plot for all these regions, which would be the track time distribution. This is to check if the low tan lambda and low momentum
region are accidental tracks. Showing plots you expect to be fine and are fine are reassuring but helpful to show to our collaborators
and can at least help us understand what exactly the selection is.
Did you double check if the plot you said was a fiducial cluster selection but did not appear to be so during the meeting as you said you would?
You never got back to me offline about that question I had.
These are not all my questions about your plots, and I hope you are willing to look past whatever issue you have with me and find a way to
communicate more offline so we can collaborate effectively to successfully analyze this data together and produce a result in a timely manner.
I have created a confluence page intended to be used as a portal for sharing information about the 2021 data analysis.
You can find it at
Please feel free to update this page with additional information.
I have added three new analyses to that page, two adding more information about the FEE analyses of runs 14168 @ 3.74GeV and 14628 @ 1.92GeV which I presented at our Wednesday meeting, and one updating an earlier analysis of the data
we accumulated using the SVT positioning wires as targets. Please let me know if you have any questions or comments regarding those analyses.
Enjoy the weekend,
Use REPLY-ALL to reply to list
To unsubscribe from the HPS-SOFTWARE list, click the following link: