Re: [BLAST_ANAWARE] reconstruction efficiencies

From: Wang Xu (wangxu@mitme7.mit.edu)
Date: Fri Jul 18 2003 - 12:28:49 EDT


Dear Chris,
        It is very nice.
        From your graph and my experience, the efficiency of newton fit
can still be improved. Because new fast fit find the tracks, but newton
fit may not. I believe efficiency of new fastfit(or fast fit)should be
same or very close.
        Wang

On Fri, 18 Jul 2003, Tancredi Botto wrote:

>
> Chris,
> I believe that blue means fast fit tracks, green means newton tracks
> (ie. not no fast fit, not no newton)
>
> It'd be very good if recon efficiencies are limited to events with less
> than a perfect hit distribution. I think it is a matter of learning how
> to deal with stubs only or even fake stubs and with crossing-cell tracks
>
> Somehow (this is limited statistics) I have the impression that things
> were not as bad for v2_15 back in march.
>
> But it should be use to plug boxes in and out of MC, which may be
> a useful debug for FF/newton. Also, a better FF may help on these
> issues
>
> --
> ________________________________________________________________________________
> Tancredi Botto, phone: +1-617-253-9204 mobile: +1-978-490-4124
> research scientist MIT/Bates, 21 Manning Av Middleton MA, 01949
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> On Fri, 18 Jul 2003, Chris Crawford wrote:
>
> > hi,
> > i went through quite a few events in nsed in an attempt to trace down
> > inefficiencies in the wire chamber reconstruction. the enclosed
> > histograms break down the misreconstructed events in one paddle
> > combination (0-14) (i also did 1-14, 1-13). the sample data set was
> > taken from runs 705-708 (first 25k events in each), and strict cuts were
> > placed on timing and coplanarity.
> > black: events w/o 3 segments; mostly no hits in entire chamber,
> > partially an issue of acceptance; found 1 tree event
> > blue: no fast fit track; mostly overloads (too many segment
> > combinations, fixable by loosening cut)
> > green: no newton fit track: mostly from 3-4 stub events, with ~8/18 hits
> > red: good events
> >
> > conclusion: the inefficiencies are mostly due to wire chamber
> > efficiency and calibration, although some improvements are possible in
> > the sofware to salvage these events. the forward events are less
> > efficient because of missing hits, miss-calibrations between adjacent
> > cells (these stubs are almost never 3 hits in one cell), and a more
> > complicated acceptance region for curved tracks (as opposed to a true
> > inefficiency).
> > some things which should be improved in software: a) make more of an
> > effort to use real hits instead of fake stubs (no hit data). this is
> > important for the newton fitter. b) smarter (dynamic) cuts so i) that
> > we don't loose real stubs/segs/tracks, and ii) we don't get
> > bloated/overloaded with too many tracks. hopefully better calibrations
> > will abate this problem, anyways. in my limited viewing of the data, i
> > did not find any obvious bugs in the reconstruction software, and almost
> > of the missed events were due to the above explanations.
> >
> > finally, if possible, i would like to take a short unpol H run at .5
> > sccm (~1/2 hr) sometime, to have some more recent data for WC
> > callibrations, and testing reconstruction efficiencies. since the runs
> > 700-800, some of the missing cells have been restored. i would like to
> > see how this improves our current reconstruction.
> > --chris
> >
>



This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST