Project

General

Profile

Tracker Software Review: Meeting 2

May 26, 2017: 1400 BST

Agenda

  1. Introduction - K. Long
  2. Actions from previous meeting
    1. Compare data/MC efficiencies for 8681
    2. In adding unused spacepoint, show the distribution of distance of unused s.p & its correlation with #pe
    3. Low pt: Consider doing a straight track fit first and pass to Kalman if it passes chi2/other requirements
    4. Generate list of uncertainties and show what has been done so far (alignment, scattering, energy loss, field uniformity, etc)
    5. Check dead channel & noise handling in MC
  3. Pattern recognition update: A. Dobbs tracker-recon-status.pdf, p-value, 8681, CR p-value.pdf
  4. Kalman & track fitting overview: C. Hunt Track_fit_overview.pdf
  5. Next Steps for review
  6. AoB

Additional documentation

  • Tracker Software Paper: attachment:"https://micewww.pp.rl.ac.uk/attachments/download/8741/MICE-Tracker-Software-Publication.pdf"

Minutes from previous meeting (April 27, 2017)

  • Attendance: KL, PK, DR, CR, DC, AD, MU, EO, AB
  • Algorithm:
    DC: In determining #turns, do we have enough information to tell if there could be > 1 rotation between stations? Has such a restriction been imposed — can it be imposed — in selecting “ideal events”?
  • Efficiency:
    DC: Does MC show the same pattern of lower TKD efficiency?
    — are dead channels and noise handled correctly in the MC?
    MU: Noise smearing added by Heidt; AD notes it’s not part of standard production
    KL: we don’t want to rely on the MC for this unless we really have to
    Action: Check dead channel & noise handling in MC
    Action: Check MC of this run 8681 & compare efficiencies with data
  • Sources of error:
    CR: Need a list of sources of uncertainty: alignment, scattering, energy loss, field uniformity, etc and present what studies have been done — Action
  • Unused spacepoints:
    DC: In adding unused spacepoint, show the distribution of distance of unused s.p & its correlation with #pe — Action
  • Pt singularity:
    AB, DC: why not just run straight track fit 1st, if it satisfies chi2, then send to Kalman instead of trying to do a helical?
    Q: can Kalman reconstruct a helical from the straight?
    CH: Think so, if seeded properly.
    Action: This should be studied

Dial-in information

http://mice.iit.edu/phonebridge.html


tracker-recon-status.pdf (1.43 MB) Dobbs, Adam, 25 May 2017 17:49

Track_fit_overview.pdf (161 KB) Hunt, Christopher, 26 May 2017 14:04

p-value.pdf (14.1 KB) Rogers, Chris, 26 May 2017 15:18

Tracker-Sw-Mtg2-Actions.pdf (40.4 KB) Rajaram, Durga, 09 June 2017 03:14