[Tpe] tracking problems in TPE

[Mac Mestayer]

Hello all;

I've been talking ot Rob and Stepan about the strange changes
to tracking when one artificially changes the assigned torus
polarity.  Here are some thoughts on things to look at.

We all agree that more information is needed.  Rob will
re-cook some data and write out the TBLA bank as well as
the DC bank and others.  Once that is done, we will be able
to look at events with ced.  Perhaps these events will
have nearby accidental groups of hits which might be
selected instead of the true one.

Some of the problems may just be software bugs; for example,
bad pointers or perhaps different road files for positive and
negative tracks instead of for outbending and inbending tracks.

Some of the problems may be due to unreasonably large chi-squared
cuts (> 100/d.f. ?)

Some of the problems might be due to out-of-time accidental tracks.
These will have very unusual velocities: perhaps very low beta or
beta greater than one.  The drift chamber residuals will also show
a strange pattern with a mean time residual not centered at zero
but off-center.

By the way, once the TBLA bank is written out the TRKMON program
can be run on the data and turn the bank into an ntuple which
can be studied with PAW.

 				- Mac

"mestayer at jlab.org", (757)-269-7252