Privacy and Security Notice

Archived Messages for CLAS_DRIFT_CHAMBERS_1997@cebaf.gov: bug in trk_fill_bht.F

bug in trk_fill_bht.F

James A. Mueller (mueller@blakey.phyast.pitt.edu)
Sat, 04 Oct 1997 12:04:36 -0400 (EDT)

While trying to understand some strange effects presented by
Liming at the CLACOM meeting yesterday, I found a bug in the ana routine
that makes the hit-based-tracking n-tuples from the recsis banks.
Specifically, there was a error associating tracks in the HBTR and DCLA
bank. This results in an error in the information in ntupole 112 in the
/trk/hbt subdirectory of your hbook file in the following manner.

Consider an elastic event with an electron in sector one and
a proton in sector 4.
a) The entry for the lowest sector track (in this case e) will be
correct
b) The entry for the higher sector track (in this case p) will have
hit-level information (sector, number of wires hit, which wires
hit) correct, but ...
c) the track based information (q, p, theta, phi, etc) will be from
the lower sector track. (In this case it will look like you have
negative protons at the wrong theta and phi.)

I want to stress that this is only a problem with this ntuple, and NOT
WITH THE OUTPUT BANKS themselves. However, I know that Rob Feuerbach's
analysis of the DC HV scan was based on this ntuple, and the plots
presented yesterday by Liming suggest that a similar error may be
effecting his code. I have corrected this bug and committed it to the
CVS repository. Anyone who uses histograms made by user_ana in the
trk/hbt directory, should probably checkout the ana area and rebuild.
If you don't use these histograms, you can ignore this.

-Jim Mueller