Hi BRATs,
I've started a new development cycle of BRAT.
Version: 1.16.0
CVS Tag: BRAT-1-16-0
Changes:
* BrMultRdo is now compiled but still not used.
* Introduced Mult base class BrMultModule for almost all TMA and SMA
modules. This class has the (protected) method FindVertex, so that
all modules deriving from this class will use the same method to
obtain the vertex position (in reference to Steve's previous
concern). This class (or a derived class) could be expanded, to
hold many of the almost identical TMA and SMA algorithms.
* Important bug fixes to Br[Tile|Si|Mult]CentModules. In particular,
they will not overwrite each others histgrams.
* Important bug fix to Br[Si|Tile]Calibration when returning
centrality functions. These were introduced by me in BRAT 1.15.9.
* Preliminary code to use total deposited energy (rather than the
multiplicity) for the centrality calibrations have been put into
Br[Tile|Si]CentCalModule. Br[Tile|Si]CentModule can not use this as
of now.
I've checked the calibrations for the centrality as determind from
Br[Tile|Si|Mult]CentModules, and they all seem ok in |Vz| < 20cm,
while Br[Tile|Si]CentModules does reasonably well up to |Vz| < 30cm,
and with a bit of good will, up to |Vz| < 35cm.
I've attached a plot "cmpPlots.ps" to show you this. The histograms
are normalised to "minimum-bias". That is trigger 4, |Vz| < 50cm, and
for each module
* BrMultCentModule: at least one hit in TMA above threshold AND
at least one hit in SMA above threshold
* BrSiCentModule: at least one hit in SMA above threshold
* BrTileCentModule: at least one hit in TMA above threshold
The plots are from run 2481. There's approx 300,000 events that go
into these plots.
The centrality calibrations was done using the same event sample, and
for both the SMA and TMA Rdo code, the outlier correction was set to
kNoCorrection. That means, that to reliable use the
Br[Tile|Si]CentModules, you _must_ do no outlier correction in
Br[Tile|Si]RdoModule. I believe the same goes for BrMultCentModule.
It seems that there's a systematic rise around 10cm < Vz < 30cm for
the TMA and SMA code. I believe that can be remedied by using a 0'th
degree polynomial parameterisation (rather than a 1'st as presently
done), but I haven't checked it. See also the attached file
"cutPlots.ps"
Yours,
Christian -----------------------------------------------------------
Holm Christensen Phone: (+45) 35 35 96 91
Sankt Hansgade 23, 1. th. Office: (+45) 353 25 305
DK-2200 Copenhagen N Web: www.nbi.dk/~cholm
Denmark Email: cholm@nbi.dk
This archive was generated by hypermail 2b29 : Wed May 23 2001 - 08:55:19 EDT