From Francois Roudot, to all offline users and drift chamber list.
As we discuss at the last offline meeting on wednesday, I have set a
new cvs package into the tree for histogramming and analysis. The name
of this package is ana. I have created also a new version of the user
package : user_ana.
About ana package:
-----------------
The content of this package follows:
CVS/ ec_book.F seb_fill.F trk_fill.F
Makefile ec_fill.F tag_book.F trk_fill_hbt.F
cc_book.F histtcl.inc tag_fill.F trk_fill_raw.F
cc_fill.F histtcl_init.F trk_book.F trk_fill_tbt.F
depends/ sc_book.F trk_book_hbt.F user_book.F
ec1_book.F sc_fill.F trk_book_raw.F user_fill.F
ec1_fill.F seb_book.F trk_book_tbt.F
It is a bunch of fortran routine defining the filling and the booking
of histograms for each packages (cc, ec, ec1, sc, seb, tag, trk and
user).
As of now, the package_book.F and package_fill.F routines are dummy
routines, execpt for the trk package.
+ The purpose of the package_book.F routines is to book histograms and
ntuples for the given package in his own //LUN1/PACKAGE paw
subdirectory in the re****.rzn output file.
The name of the paw subdirectories are: /CC, /EC1, /EC, /SC,
/SEB, /TAG, /TRK and /USER.
+ The purpose of the package_fill.F routines is to fill histograms and
ntuples for the given package in the previous package define paw
directory in the re****.rzn output file.
+ In addition to that, as you can see below, the routine histtcl.F set
all the tcl variable to control histograms and ntuple booking and
filling to true:
LTRK_H_DO TRUE -> perform tracking histo
LEGN_H_DO TRUE -> perform calorimeter reconstruction histo
LCC_H_DO TRUE -> perform cherenkov reconstruction histo
LTOF_H_DO TRUE -> perform TOF reconstruction histo
LST_H_DO TRUE -> perform start counter reconstruction histo
LUSR0_H_DO FALSE -> perform USER code BEFORE event processing histo
LUSR1_H_DO TRUE -> perform USER code after EVENT processing histo
LUSR2_H_DO FALSE -> extra flag
LTAGGER_H_DO TRUE -> perform tagger stuff histo
LEC1_H_DO TRUE -> perform large angle calorimeter histo
LSEB_H_DO TRUE -> perform Simple Event Builder histo
These variables are defined in the histtcl.inc file.
+ As of now only trk_book (which calls trk_book_raw, trk_book_hbt and
trk_book_tbt) is not a dummy booking routine. This routine create 3
subdirectories into the //LUN1/TRK (/RAW, /HBT and /TBT) and then
book a bunch of histograms in these subdirectories.
+ The trk_fill (which calls trk_fill_raw, trk_fill_hbt and
trk_fill_tbt) then fill the defined histograms and ntuple in the
proper paw subdirectory.
NB: The makefile scheme has been modified in order to check all
include files in all packages.
The purpose of this new package is to have a common place for
histograms ntuple and other analyis stuff for the detector packages.
I strongly recommand other package responsible to use this new
poackage in order to have a bunch of well define histograms for the
upcomming June run.
I have tested this booking and filling structure and it works. All the
work to book and fill histograms for other packages is already
prepare, you juste have to place your hbook and hfill call in your
package routine...
About user_ana package:
----------------------
In addition to this new package I have had to create a new version
ofthe user package (user_ana) which is linked to the ana library via
his own Makefile.
In this version on user_ana, 3 routines are no longer dummy.
+ user_tcl_init calls histtcl where the histogramming tcl variable are
defined.
+ user_init calls the packages booking routines if the packages are
turned on and if the histograming/analysis is set to true for this
package.
+ user_evnt call the packages filling routines under the same
conditions.
In addition , the makefile is modified to link the code with the ana
library.
The newint.tcl file is also modified to show a way to control the
histogramming/analysis tcl variables.
General remarks:
---------------
The advantage of this scheme is that the size of RECSIS for common
(Analysis) use will not increased.
It seems important to me that we will have a standard set of
histograms including all packages to be abble to make diagnostics as
quickly as possible for the entire detector during the upcomming run.
That's why I strongly recommand do define histograms with this new
package.
It will be great also if all package histogram responsible provide a
bunch of kumac file in order to made the detector diagnostic.
To get started with this histogram/analysis package you have only to
cvs checkout user_ana and ana. That should be sufficient...
I hope this description of this new package is complete and
understandable, if you have any problems and/or questions, feel free
to contact me.
Francois.
--
Francois Roudot : Roudot@cebaf.gov
CEBAF : 12000 Jefferson Avenue Home : 1634 E. OceanView Ave. #3D
Newport News VA-23606 Norfolk VA-23503
Off. (757) 269-7680 Pho. (757) 587-4855
Fax. (757) 269-6273