Status of Histogramming Track Finder

Report on the general tracking / vertexing meeting 7/26/01

• htf is linked to the gtr executable and ready to be used
• recent modifications include:
  • content id of GTracks produced by htf can be set in rcp
  • no debugging printouts
  • the timing information plot can be switched on/off in rcp
• looking for the track quality and ghosts
• timing issues

The analysis on the percentage of ghost/fake tracks is in progress. The results obtained with htf and trf++ for a sample of Z to mumu with 4 MB for the reco version t01.52.00 suggest that the reconstruction efficiency for primary tracks above 1 GeV demonstrated by htf is about 90 percent versus 80 percent shown by trf++, and the fake rate is 2-3 times higher. The fake rate can be reduced by imposing higher chi2 cut, however more sophysticated solutions are being investigated.

htf is 2-3 times slower than the recent version of trf++ This is going to improve with a new version of htf which is coming.