Project

General

Profile

Actions

PC03022016

Wednesday 3 February 2016 at 15:00 GMT
Room CR09 RAL

Summary of Actions

Task Person Added Issue Number Comment
Fix Pattern Recognition bugs that cause loss in efficiency A. Dobbs 06/01/16 #1807
GRID MC running D. Maletic, D. Rajaram 02/12/15 Needs a champion - D. Maletic? Local running under DR until then.
Fix tracker coordinate system flips Tracker group 18/11/2015 #1792 Mostly done now - C. Hunt should have fixed by next meeting.
Add data_analyzable flag for each detector in post processing to CDB Detector software heads 16/09/2015 #1585
Add tracker efficiency plots to MAUS P. Kyberd 16/09/2015 -
Optimize residuals for final tracks C. Hunt - - Done for straights
Fix close to paraxial tracks producing nan output C. Hunt - -
Update tracker online plots M. Uchida - -
Detectors to provide uncertainties from reconstructions on: positions; momenta; tof time A. Dobbs - -
Implement dE/dx in Runge-Kutta J. Greis - - Branch to be ready for merging by next meeting
Verify scifi geometry correct in CDB C. Heidt, P. Kyberd 30/09/2015 - Check by running with MAUS (use ID 671)
Generate G4BL library R. Bayes 30/09/2015 -
A. Dobbs to kick off discussion on online EV with Y. Karadzhov 2016-01-20 - Do before next meeting

Release status

  • Current release = 1.4.0
  • Targets for release 1.5.0:
    • Fix tracker MC flip issues
    • Optimised tracker helical pattern recognition
    • Geometry update to support opera field model (#1810)
  • Targets for future release:
    • Trigger MC
    • Tracker -- Kalman updates

Schedule

Detectors

  • Detector integration (D. Rajaram)
    • ACTION: Add data_analyzable flag for each detector in post processing to CDB (Detector software heads) #1585, #1689
  • Tracker (P. Kyberd)
    • ACTION Fix Pattern Recognition bugs that cause loss in efficiency #1807 (A. Dobbs)
    • ACTION Fix tracker coordinate system flips
    • MC studies of effect of 1T field on tracker performance (P. Kyberd, Analysis Group)
    • Space-point, track-finding efficiency studies partly blocked by lack of trigger #1182
    • ACTION: Optimize residuals for final tracks (C. Hunt)
    • ACTION: Fix close to paraxial tracks producing nan output (C. Hunt)
    • ACTION: Update tracker online plots (M. Uchida)
    • Flag low pt for downstream tracker reconstruction? (A. Dobbs)
  • TOF (D. Rajaram)
    • Restructure TOF MC Digitizer - part of trigger MC (D. Rajaram)
  • EMR (F. Drielsma)
    • attachment:maus151216_fd.pdf
  • KL (M. Bogomilov)
  • CKOV (A. Liu, D. Rajaram)
    • Calibration
    • Validate CKOV MC against data

Global Reconstruction

  • Tracking (J. Greis)
  • PID (C. Pidcott)
  • Detector output
    • Get detector output in global coordinate system (C. Heidt, C. Hunt), see #1631
  • ACTION: Detectors to provide uncertainties from reconstructions on: positions; momenta; tof time (A. Dobbs)
  • ACTION: Implement dE/dx in Runge-Kutta (J. Greis)
    • JG proposing merge into trunk. AD to get it in

Monte Carlo

  • Monte Carlo trigger (D. Rajaram) #1182
    • Restructure TOF Digitisation and incorporate MapCppTrigger (D. Rajaram)

Geometry

  • Handling field maps in MAUS (#1810)

G4BeamLine

Online Reconstruction

Batch Offline Reconstruction

  • MAUS running fails on GRID (#1815)
  • GRID running 1.0.0
    • Enable MC GRID running again #1739
    • execute_against_data.py to always produce tarball Is this still relevant?
  • Offline Rec running 1.4.0

Infrastructure

  • InputCppRoot still using JSON (and so super slow) (D. Rajaram) (#1814)
  • Fix geometry test which randomly produces large SOAP error (R. Bayes)
  • Fix problems when setting header_and_footer_mode to "dont_append" (C. Rogers, C. Pidcott)
  • Fix matplotlib install (A. Dobbs)

Performance

  • Speed improvements for MC
  • Beam generation excessive resource usage
    • Query analysis group requirements (in particular how many G4BL beams we need)
    • Consider alternative upstream beam generation code

Tools

Event Viewer

  • Add event viewer to MAUS #1799
  • * ACTION* A. Dobbs to kick off discussion on online EV with Y. Karadzhov before next meeting

Documentation

  • Update tracker (C. Heidt)
  • Update geometry (R. Bayes)

AOB

International Toll-Free Dial-in numbers

  • Switzerland: 0800705351
  • UK: 08004960576
  • US: 1-866-740-1260
  • Other numbers
  • Access Code: 5673390

Notes

Attendance

M. Savic, C. Pidcott, J. Greis, M. Uchida, F. Drielsma, C. Heidt, R. Bayes, D. Rajaram

Release Status

A. Dobbs will release 1.5.0 when tracker group provides MC fixes

Schedule

Detectors

  • All detectors
    • ACTION: Urgently need validation plot from detectors to integrate with geometry & reconstruction to make sure geometry+reco is OK
  • KL
  • Ckov
  • EMR
  • Tracker
    • Star problem (fiber lengths in geometry), large residuals between reco and MC truth.
    • Sliding residuals is the remaining issue.
      • C. Hunt has pushed some fixes - now in MAUS. Heidt working on geometry solution for other half of the problem.
    • M. Uchida expecting to get tracker online plots updated in next few weeks
      • These are efficiency and validation plots, not intended for online reco but DR thinks it's worth maybe having some subset of it for online reco
  • TOF

Global Reconstruction

  • Global team will try to provide a branch for merging in their code before the next meeting.
    • J. Greis - testing of global tracks is blocked by tracker MC. But ready to push code de/dx into MAUS
      • built on Jenkins, but seeing some application test failures on private machine. Will create issue and if MAUS man is happy JG will propose a merge

Monte Carlo

Geometry

  • RB waiting for alignment corrections from FD before releasing new geometry
  • CHeidt asked RB to check with him before final release of geo, since tracker group may have some more fixes.
  • We need detector validation to integrate with geometry to ensure that the geometry representation is OK for MC and data reconstruction (see above)

G4BeamLine

RB will provide a link to non-grid-hosted G4BL chunks for people interested in testing

Online Reconstruction

Batch Offline Reconstruction

  • DR waiting for tracker fixes and new release of MAUS before proceeding with MC & reprocessing on MLCR
  • D. Rajaram will continue working with D. Maletic to get full GRID MC back up again.
    • DM getting up to speed, getting certificates etc to poke around on GRID.

Infrastructure

  • MAUS running fails on GRID. Portability issue related to numpy build. Issue created.
  • Reco quality: JM creating a table for reco quality flags (~2 weeks), DR working on integrating TOF quality, will test with table on preprod

Performance

MC speed needs to be addressed when we find time.

Tools

AOB

  • Software workshop at CM44 likely to be on the Friday afternoon, A. Dobbs to confirm

Next meeting: 15:00 GMT, 17th Feb 16

Updated by Rajaram, Durga over 5 years ago · 8 revisions