BlastLib2 and lrd modified with the following changes:
1. lrd now takes an "Outdir" where it would dump flr and ep_skim ntuples.
When Outdir is set, flr will be recreated in this dir without the old
ntuple being backed up in the root file.
DST tree is backed up however. In fact lrd would copied the original dst
file into that dir and operate on the copied file. still have trouble
writing to a second DST while reading the original one.
So one gets a fat dst file but clean flr. the copy is handled by lrd
within, no more operations are needed other than "lrd ####".
2. Unless it sees lr files in current dir or "Outdir" is not set, lrd
will not produce lr ntuple at all. Because in these cases, no-track
events are not known to lrd in any sense so it is not able to produce lr
ntuple any different than the flr.
3. Problem of not being able to improve proton resolution as lrn did is
fixed. recrunch with lrd and proper Wire.Cal will give 20MeV proton
momentum resolution.
4. BATS has never been implemented in DST correctly prior to last couple
days!!
Before v3_3 is used, there simply is no bats entries in DST. noticed
couple days ago that bats codes in v3_3 pertain to DST is no more than a
copy of TOF codes so in DST crunched since run 10005, one get 0-16 for
"fBat.fPad".
These problems are corrected. But
ALL DST SO FAR HAVE NO OR WRONG BATS INFO!!!
This means recrunch from dst would cause loss of BATS information. Not a
concern for D2 runs. Program will not seg-fault because of missing bats.
How about H2 runs?
5. two dozen new entries added to lr/flr ntuples including geometric
average of ADC's, neutron counter ADC's linked to tracks. see lrn_event.h
for detail.
Because L20l/r L15l/r are all on the right side, it is a bit hard to
associate hits to the tracks in the ntuple, especially when both
sides have tracks. still the best way is going into the DST.
Chi
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:31 EST