CM43 Software Workshop¶

Date: Wednesday 28th October 2015
Time: 10:00am - 5:30pm
Venue: CR09, ATLAS building, RAL
Attendance: A. Dobbs, F. Drielsma, J. Greis, C. Hunt, P. Kyberd, A. Liu, S. Middleton, T. Mohayai, J. Nugent, C. Pidcott, D. Rajaram, M. Savic, C. Rogers, M. Uchida, A. Young

Rough Agenda¶

Morning Session (10:00am - 12:45pm)¶

10:00am Opening Discussion

10:45am MAUS Development (A. Dobbs)

11:30am Coffee

11:45am Step IV Discussion and Open Issues

12:45pm Lunch

Afternoon Session (1:45pm - 5:30pm)¶

1:45pm Event Viewer Discussion

http://micewww.pp.rl.ac.uk/projects/maus/issues

3:00pm Coffee

3:15pm Problem Solving Q&A

Notes¶

CKOV¶

• CKOV no longer works with Stage4.dat legacy geometry - need to disable CKOV modules in top level maus python script when running a simulation with this geometry (other get the GEANT not ready error) #1781

EMR¶

• Change to CDB causing faults with current release unit tests. F. Drielsma working on it

Tracker¶

• x, y, px, py flips observed in final scifi track output bu Global Tracking in both helical and straight tracks - interfering with current Global development efforts. May be partly due to old TOF geometry being used (J. Greis to check his work)
• Similar issues seen when visualising track points with event viewer
• pz recon is very poor #1782
• C. Hunt aware of it all - working on it
• J. Nugent seeing good result from millipede on tracker alignment, expect for stations 4 and 5 in downstream tracker which are very poor. This in contrast to C. Hunt's alignment work, which produces good results only. Cause as yet unknown.

Event Viewer¶

• Lots of progress from M. Savic
• Consider using alternative existing solutions, such as the CMS iSpy app, or HepRApp
• Functional system implemented using existing event viewer and HepRApp
• C. Rogers keen to get in MLCR as soon as possible
• M. Savic now set up on Launchpad, event viewer repository at: https://code.launchpad.net/mice-event-viewer

MC¶

• A. Liu noticed MC very very slow when pushing lots of particles through tracker - code needs profiling (A. Dobbs) #1785

Infrastructure¶

• Y. Karadzhov looking into source of memory leaks, so found if final scifi tracks not added to scifi event memory leaks disappear - but destructors look good (investigations continuing)