_ Recon efficiency - WITH tof cuts
For runs 3063-3098 ("golden runs") the lazy fitter
reconstructs above 90 % of supposedly "good events"
However fast or newton tracking eff. can still be as
low as 50 % depending on sector/particle/angle. Note:
there were also problems with 1 bad box/sector
Fast or newton seem to reconstruct a very high percentage
of all those events with 3 segments/sector
But quite a lot of the good events do not recontruct up to 3
segments.
Above considerations are for runs where the Wch-TDC S/N is 20-40
For "abs runs" late in december we have S/N 1-2. There also
lazy tracker is quite inefficient (60-70%) and fast/newton
are much poorer as well
_ Tracking Development:
Chi has been testing a "fake stub" algorithm that allows him to
recover most events with only two segments and 0/1 stubs
The fake stub can be taken from the TOFs out-of-plane angle.
Note: the algorythm will be expanded to include also single hits,
not just stubs.
Deadline for CVS submission is tomorrow at midnight. A new tag
will be made. And a new re-iteration of results..
- Other development:
It is painfully obvious to see that the recon script will hang
for several minutes on single events for whihc we generate way
too many track candidates. Sometimes there are several 1000's
candidates. Corrections are possible to discard these few annoying
events at the tracking and cluster finding step. Will be in the
next cvs release.
Drift time calculation (garfield) for a lorentz angle have been
completed. Results need to be parametrized for BLAST use.
The new drift time tables will be wch-cell-dependent and angle
dependent which will make them more difficult to use (also, will have to
be used iteratively). Results ready in a few days.
Wong looking at cluster find algorythm also in view of above changes.
- Other:
Now there is general consensus not to go to V3 reconstruction
before the run.
________________________________________________________________________________
Tancredi Botto, phone: +1-617-253-9204 mobile: +1-978-490-4124
research scientist MIT/Bates, 21 Manning Av Middleton MA, 01949
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
---------- Forwarded message ----------
Date: Tue, 18 Mar 2003 14:05:52 -0500 (EST)
From: Tancredi Botto <tancredi@mitlns.mit.edu>
Analysis Meeting wendsday 10 AM, Bates conference room x3,9296
please read on
_ Tracking:
*efficiency*
2/3 segment tracking
stub/segments stats
status of newton/fast/azy
*resolution* garfield/drift-time relationship
wire/tdc offsets
_ General:
*cvs updates*
*should we start migrating to v3 NOW after all..*
_ Set a suitable data for a db discussion: adrian/jason/chris/nikolai
_ Montecarlo studies..
-- ________________________________________________________________________________ Tancredi Botto, phone: +1-617-253-9204 mobile: +1-978-490-4124 research scientist MIT/Bates, 21 Manning Av Middleton MA, 01949 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST