For the efficiency studies on real data, several runs (14577, 16784, 17137)
were reprocessed using the three different pattern recognition packages in the tracking.
The result of these studies is that CATS finds 10-18% more
s than RANGER and TEMA,
depending on the applied cuts. CATS also produces better signal to background ratios and
significantly narrower
-peaks. In additional CATS has demonstrated higher efficiency
for
signal, it finds 20% more
s than RANGER and TEMA.
It can be seen that results obtained on MC and real data are in the good
agreement.
Since June 2001 CATS is the default track recognition algorithm for
the Pattern Tracker.
It has been used to reprocess all data collected in 2000 and 2002/2003.
The relatively high efficiency of CATS for steep tracks crossing different
sectors of the detector is especially useful for alignment procedures.
In addition, the space-points produced by the package are used for the initial
alignment of the OTR superlayers and cross-checks of the relative positions
of ITR modules and OTR superlayers.
Yury Gorbunov
2010-10-21