Difference between revisions of "AIEC Weekly Meeting Nov. 04, 2022"

From epsciwiki
Jump to navigation Jump to search
(Created page with "=== Agenda: === # Previous Meeting #: # Activities: #* [https://halldweb.jlab.org/grafana/dashboard/db/aiec-cdc?refresh=15m&orgId=1 Grafan...")
 
 
Line 1: Line 1:
 
=== Agenda: ===
 
=== Agenda: ===
−
# [[AIEC Weekly Meeting Sep. 8 2022| Previous Meeting]]
+
# [[AIEC Weekly Meeting Sep. 29 2022| Previous Meeting]]
 
#:
 
#:
 
# Activities:
 
# Activities:
−
#* [https://halldweb.jlab.org/grafana/dashboard/db/aiec-cdc?refresh=15m&orgId=1 Grafana] Update:
+
#* NeurIPs Poster:
−
#** Diana working on fixing data access issues.
+
#** Diana to make changes and send out for another round of review
−
#* Torri working on EPICS, RCDB, CCDB documentation for others, Naomi working on getting specific documentation on tables in the CCDB
 
−
#* Diana/Thomas working on plot look books to make model evaluation and behavior easier to understand, with plans to turn control back on maybe next Thursday pending Naomi's sign off
 
−
#* Neurips ML for Physical Sciences paper will be submitted tonight
 
−
#* Naomi suggested submitting an abstract (due 12/14/22 at 7:59 PM ET) for [https://indico.cern.ch/event/1158681/abstracts/ HEP2023] in January
 
−
#* Gathering some preliminary data from Primex corresponding to extreme low pressure, low temperature runs that are out of our training domain
 
 
#
 
#
−
#* CPP
+
#* [https://www.overleaf.com/read/zrnvkfjvyrnz | NeurIPS paper]  
−
#** CPP Calibrations (Naomi)
 
−
#
 
−
#* [https://halldweb.jlab.org/DocDB/0057/005706/004/CDC_gain_dependence_of_PID.pdf Quantification of bad gain calibrations]
 
−
#** Looking at initial results comparing traditional GCF compared to roboCDC GCF in progress
 
−
#** HVB Currents during initialization are not consistent with what we see during actual running, but are consistent with data from the HVBs from EPICS
 
−
#** Naomi identified why our model never learned anything useful from HVB current, related to the CDC alcohol swap between 2018 and 2020 run periods. This affects the gains. See also the data in these plots [https://halldweb.jlab.org/DocDB/0057/005715/001/GCF_201820202021.html GCF_gasdensity] that Diana's model used for training. GCFs vs gas density for 2018 and 2020 should not overlap.
 
−
#** [https://halldweb.jlab.org/wiki/index.php/CPP_DAQ/Monitoring_Test_Spring_2022#AIEC_planning CPP plans]
 
−
#** [https://halldweb.jlab.org/wiki/index.php/CPP_DAQ/Monitoring_Test_Spring_2022 CPP preparations]
 
 
#
 
#
 
#* Capstone Students
 
#* Capstone Students
−
#** Prepare for Oct 13th presentation to capstone student candidates.
+
#** Thomas to send meeting request for next Thursday
−
#** Oct. - Dec. will purely be EDA and basic project understanding
+
#** Send students 2018, 2020, and 2021 data for CDC to get familiar with the data
−
#*** 4 - 6+ hours of meetings to be expected in this timeframe
+
#** Also get correction for 2018 data, then send alcohol-corrected data to students
−
#** January will really get started.
+
#** Get started on FCAL data
−
#** Diana (and Torri?): Proof of concept retrieval of all calibrations, EPICs data, RCBD(?) data for a single run in preparation for when capstone students get going.
 
−
#** Is there any need to alert MIS that we expect new User logins to be needed by end of October?
 
−
#** What will they need access to?
 
−
#*** JupyterHub resources?
 
−
#*** ssh login to farm?
 
−
#*** AIEC GitHub repo?
 
−
#*** anything else?
 
 
#
 
#
 
#* FDC calibration
 
#* FDC calibration
Line 56: Line 36:
 
# Talks/Presentations
 
# Talks/Presentations
 
#:
 
#:
−
#* Torri and Diana presenting at QNP [https://indico.jlab.org/event/344/overview (QNP)]
+
#* Diana at NeurIPS (In-Person: November 28th - Dec 4th, Attending virtual workshops: Dec5th
−
#* Diana presenting at ACAT2022 (15 min talk + 5 min questions)
 
 
#:
 
#:
 
# Publications:
 
# Publications:
−
#* ACAT 2022 - abstract Done. Travel approvals in process
+
#* ACAT 2022 - NEED to write the paper - due in January.
 
#* AI4EIC Proceedings accepted [https://www.overleaf.com/project/6185656923f59237c9c3329b AI@EIC Overleaf]
 
#* AI4EIC Proceedings accepted [https://www.overleaf.com/project/6185656923f59237c9c3329b AI@EIC Overleaf]
 
#* ACAT 2021 Proceedings submitted [https://www.overleaf.com/project/61a92b255c1b91b059016e9c ACAT Overleaf]
 
#* ACAT 2021 Proceedings submitted [https://www.overleaf.com/project/61a92b255c1b91b059016e9c ACAT Overleaf]
 
#* NIM paper in progress [https://www.overleaf.com/project/61769c2c9b165d827b8c9e59 NIM Overleaf]
 
#* NIM paper in progress [https://www.overleaf.com/project/61769c2c9b165d827b8c9e59 NIM Overleaf]

Latest revision as of 18:58, 4 November 2022

Agenda:

  1. Previous Meeting
  2. Activities:
    • NeurIPs Poster:
      • Diana to make changes and send out for another round of review
    • Capstone Students
      • Thomas to send meeting request for next Thursday
      • Send students 2018, 2020, and 2021 data for CDC to get familiar with the data
      • Also get correction for 2018 data, then send alcohol-corrected data to students
      • Get started on FCAL data
    • FDC calibration
      • Reach out to Lubomir alternative FDC for cosmics- David.
      • 2020 CDC-FDC model results: CDC-FDC model slides
      • Potentially give capstone students FDC calibration model development
      • Discuss with Simon how useful FDC calibration constants are for physics analyses
    • Next steps
      • CLAS12 Polarized Targets
        • Update from James (8/25): concern about poor fits that give polarization (i.e. Diana's eventual target values). Can begin initial data exploration whenever we have time.
        • data from fits to obtain polarization AND relevant EPICS data are stored in json files located in /group/poltar/HallB/RGC and should be readable
      • **AIEC Auto Control Wish List:**
      • GlueX TOF
      • BCAL baseline (slides) - David follows up:
        • Do we need another sensor?
        • What are the limitations on existing sensor and logging of temp value?
  7. Workshops
  8. Talks/Presentations
    • Diana at NeurIPS (In-Person: November 28th - Dec 4th, Attending virtual workshops: Dec5th
  9. Publications:
Retrieved from "https://wiki.jlab.org/epsciwiki/index.php?title=AIEC_Weekly_Meeting_Nov._04,_2022&oldid=2601"