[BLAST_ANAWARE] dst/crunch update I: the run/fill info ntuple "chg"

From: zhangchi (zhangchi@general.lns.mit.edu)
Date: Sat Jan 03 2004 - 15:42:45 EST


Hi all,

I just checked in a dst.readme to BlastLib2. it contains information about
the new features the software group put into the new library and lrn. We
expect that the new library will be used for crunching when we come back
on line.

since some of the information is important and the readme file turns out
a bit long(300 lines), I will send out parts of the readme file in
several emails.

1. charge ntuple "chg" in lr/flr root file

an ntuple containing run/fill information is written into both lr and
flr. the ntuple is called "chg". number of entries in the chg ntuple
equals to number of fills in the run plus one. the very first
entry(retrieved by chg->GetEntry(0)) of the ntuple contains
information about the run and the entries that following contains
information fill by fill.

entries of the ntuple are:
        "fillno:runno:targ:gas:spinth:flip:fntrig:fnev:"
        "bgpp:bgpm:bgmp:bgmm:bgp0:bgm0:"
        "scpp:scpm:scmp:scmm:scp0:scm0:"
        "bgp0p:bgm0p:scp0p:scm0p:"
        "bgpt:bgmt:scpt:scmt:"
        "bgrt1:rt1:wcrt1:tkrt1:"
        "blastpol:flow:ligit:cellt:sexpol:ctube:dtube:livetime:"
        "bqmt:bqml:bqmb:bqmr:closed"
where:
 the first row is general info of the run/fill:
   // fill number, run number,
   // target(-1:empty, 0:unpol, 1:abs),
   // gas(1:H, 2:D)
   // spin angle(in deg)
   // spin flipper (0:off, 1:on)
   // fntrig: total number of triggers in the run
   // fnev: total number of events with tracks in the run

 second and third row contain beamgated(bg) and scaler(sc) charge
   conformed to charge.C convention:
   // for H2: in accordance with charge.C convetion:
   // (+/+ +/- -/+ -/-) (+/u -/u)
   // for D2: vector pols are:
   // (+/+ +/- -/+ -/-) (+/t- -/t-)
   read the entry names as: bg for gated, sc for not gated scaler
   p for + and m for -, first p/m is for beam hel, second p/m is for
   target pol(vector pol in case of D2)

   there is one exception where it is different from charge.C that bgp0
   and bgm0 contain the sum of charges in helicity +/- states, while in
   charge.C, these are strict unpol charges. if any unpol state is
   mixed into a polarized target run, bgm0-bgmp-bgmm will be the
   amount of charge on the unpol state(s).

 the 4th row contains charges for pure D2 tensor+ state: read the names
   as: bg/sc as before, the later 3 letters represent beam helicity,
   vector pol which is always 0, and tensor pol which is always +

 the 5th row contains charges for pure D2 tensor- state.

 in D2 case, bgp0, bgm0, scp0, scm0 contain sum of charges in all
  target states with + or - beam helicity.

 for unpol/empty target, all charge entries would be 0 except for
  bgp0, bgm0, scp0, scm0

 the 6th row: "bgrt1:rt1:wcrt1:tkrt1:" is event rates:
  they are all event rates with trigger type one in unit of counts/C.

  from left to right:
   beam gated scaler rate,
   event rate by counting coda events and
         divided by run time(when 2nd level trigger is active, this number
         will be considerably lower that bgrt1)
   event rate with WC hits in at least all 3 chambers in one of the
         sectors, this represent a lose collection of events that
         program should reconstruct at least one track
   event rate with WC tracks

 the 7th row are general info of the run/fill:
  blast polarity, gas flow, ligit, cell temp, sexpol pressure, ctube,
  dtube, live time. they are averaged over all scaler/epics events.

 the 8th row are bqm rates in unit of counts/C

 the last entry "closed" is set to one when the crunch is
  completed. init.C is updated to use this entry to select runs that
  are finished crunching or the runs still going. See the next section
  of this fill, see the 4th section of this file for the open/closed
issue.

Chi



This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:30 EST