Summary notes of the Software Meeting on August 16, 2012

Tracking update - Norman

Norman has been working on updating the tracking so that the correct geometry for the magnetic field can be used without all the extra rotations. Related to this is an update on the Magnetic Field, which can now be non-uniform. He has put in a Runga-Kutta method stepper and made much progress on getting the Kalman filter working. The fitter and smoother seem to be working. He expects work will be finished by the collaboration meeting. Jeremy will implement the same Magnetic Field code in SLIC, so non-uniform fields will be available there as well. Stepan will provide a 3D magnetic field map (x,y,z,Bx,By,Bz) from the existing measured y component in (x,y,z) and (Bx,By,Bz) on a plane.

Documentation.

We had an extended discussion about the need for more documentation. There has been a tremendous amount of progress in the hps-java branch of the software. Much of this progress had fairly rapid turnover of code, which meant it was very impractical to try and keep any kind of documentation up to date. In first instance, people who have questions about the code should send emails to the HPS Software <hps-software@slac.stanford.edu> mailing list. There will be an effort to reply to such emails quickly. There will also be more documentation added to the confluence pages. The documentation falls in three categories:

  1. What drivers are used/needed for an analysis?  What do the different drivers do?
  2. What is in the lcio collections?
  3. What are the algorithms used in the analysis to get the results? (e.g. how is clustering done?)

It was pointed out that although item #3 is the least likely to be asked on the mailing list, it is likely to be the most important to have documented when there are issues with a part of the detector during the 2014 commissioning run.

Updating LCIO & LCSIM

We have until now been on a frozen version of the underlying libraries of LCIO and LCSIM. Jeremy would like to move this to the "head" the latest version. We agreed this would be a good, and Sho agreed to run some test jobs on the resulting software version to make sure everything still works the same.

Action Items:
  • Norman:  Finishes the tracking update. Presentation of results in 2 weeks.
  • Stepan:  Create a 3D magnetic field map.
  • Jeremy & Sho: Update the lcio and lcsim version to latest and test resulting code. Brief presentation in 2 weeks.
  • Sarah, Kyle, Maurik: Test the ROOT interface
  • Everyone:  Ask questions on the hps-software@slac.stanford.edu list.
Next Meeting:

Next meeting will be in 2 weeks, on August 30, 2012 at the regular meeting time: 10am Pacific/1pm Eastern.

  • No labels