Core Minutes 10/17/2006
Present: Joanne Bogart, Toby Burnett,
David Chamont, James Chiang, Seth Digel, Richard Dubois,
Warren Focke, Riccardo Giannitrapani, Berrie Giebels,
Tom Glanzman, Navid Golpayegani, Heather Kelly, Michael Kuss, Bryson Lee,
Chuck Patterson, Igor Pavlin, Leon Rochester, Tracy Usher
- Service Challenge laundry list: (Tracy)
- Different calibrations – In order to combine simulation
and recon more realistically in a single job step (not a necessity,
but preferable), the code should be
able to keep track of separate calibrations. (Leon) He has a working
example for Tracker. He offered to describe it, particularly for
the benefit of developers for CAL and ACD, in an appropriate
forum. Richard suggested next week's Core meeting.
- Support of non-floating data types – Currently the interface
merit provides to CT code only supports
floating point quantities. We need to add integers and, while
we're at it, categorical variables (strings).
- Hadronic models –
Francesco is working on it. Tracy would like to make it a
separate package à la multiple scattering
- GCR – would like a fast version of track reconstruction just
good enough to draw a trajectory through the CAL. It may be that
"Tracker moments analysis" (not yet ready for prime time) would
be sufficient. (Leon) Or one might just adjust some of the knobs
already available in the standard pattern recognition.
- Backgrounds – (Toby) Bill would like some
data to study high-energy backgrounds. He suggests we use standard background
sources, but apply a low energy cutoff of 10 GeV. He would also like to
see what happens with the alternate hadronic modeling.
- Level 0 data simulation (Heather) See her
Confluence page for the latest scoop.
- MRStudio: (Riccardo) He and Marco
verified that the core libraries and command line interface work
on both Linux and Windows, and the gui also works fine on Windows, but
is broken on Linux. They expect to have it working on both platforms
in 1-2 weeks. (Richard) Is line-mode functionality working and
documented, so that Navid can look into migrating RM to MRStudio?
(Riccardo) Checkout, configure and compile are all there. The one
additional thing which RM might need is the ability to invoke a
program (e.g. test application). There is online help (just type
MRStudio --help) and if this isn't sufficient just
ask Riccardo.
- Spacecraft geometry (Richard) GBM has a
gdml model. We would like to get it into xmlGeoDbs, but don't expect
to get it done before the end-of-November runs.
- Beamtest (Michael) The GSI run is
Nov. 18-19, precisely coinciding with an SCS power outage. Richard
will try to convince them to move it, but does not expect to succeed,
hence there will be no access to the central calibration metadata database, and
no immediate runs through the pipeline.
Prompt recon will have to use ideal calibrations or, with somewhat more
work, we could set up an alternate MySQL server locally in advance and
dump the contents of the main one into it shortly before the outage.
(Richard) Any Beamtest analysis news? (Leon) Clusters in data are wider
than for MC, but there seems to be no such effect for muons. There is
nothing to report yet — it's a work in progress. (Tracy) Riccardo Rando is
finding from BT data that the CAL clustering algorithm is not always doing a
good job at finding track directions. Tracy suspects problems with attenuation
calibrations or differences in the real attenuation along the crystal versus the
simple model in simulation/recon. [See also
Riccardo's presentation at this week's Beamtest meeting.]
- Data server back-end (Richard for David Chamont)
He has made one last tag of the backwards-compatible version. All future
work will be on the new stuff which interfaces to the new front end. This
should be ready for the 1-week test at the end of November.
J. Bogart, Last Modified:
01-Jun-2010 15:45:01 -0700
previous
