a brief summay of BRAT and MDC-I plans
Alv Kjetil , JHLee and F Videbaek had a discussion on short term goals
for
the MDC-1 activities on September 15. Kris Hagel had been consulted
earlier
in the day.
Present status was reviewed. The cpu time per event is increasing
throughout a run. This may be related to memory fragmentation, not
memory
leaks, and clones arrays that have been brought under control.
The size of output files is only about 10Mb per 500 events (at 5 deg).
This
is small compared to the MDC requirements. More data has to be added
though
from T2 match, and relevant geant traks. It is not per say bad with a
small
output size, only for the MDC-1 since neither a merge program has been
written, nor has the merge option been implemented in Tom Throwe's
scripts
Some action items
a) Check current changes back into CVS (AKH)
b) The TPC local track and T2match should be completed (KH)
c) Once this is done tag CVS version as release and use for initial
MDC
reconstruction (FV)
The next set of goals is to
a) Focus at analysis modules, and 'event extraction modules.
b) Progress discussion on data structures and track definition,
acceptance corrections.
c) Add information on MDC scripts code etc to the Web pages (AKH)
d) MDC-1 reco should be fixed for errors/ problems but no overhaul
of code
structures should be made.
Longer term i.e. after MDC (about Oct 15) and to be discussed again at
at
Collaboration meeting.
a) Improvement of DC tracking algorithms
b) Geometry classes (vectors, lines, planes/surfaces,
transforations)
c) Database access
d) PID with TOF
e) Beam -Beam counter rdo summary, trigger information
f) TPC algorithms
g) Implementation of MRS to first - vertex determination later- MRS
tracking
Other outstanding code jobs
a) Replacement of TsonataMath with Tmath and BrMath.
b) Development of scripts/programs to generate 'job files' for the
typical
analysis. The existing bramsreco.cxx is too involved for average
analysis.
One could develop tools to automatically generate such files from
simpler
text files just defining modules, order, and input output.
Please comments on this particular for the longer term goals.
-- ------------------------------------------------------------------Flemming Videbaek Physics department, Brookhaven national laboratory Upton L.I., NY11973, USA
Phone: (516)-344-4106 FAX: (516)-344-1334 secr: (516)-344-7681 e-mail: videbaek@bnl.gov
This archive was generated by hypermail 2b29 : Tue Feb 01 2000 - 20:35:04 EST