|
How to Read the DIRC Mini
This recipe demonstrates how to read the DIRC mini that is being
produced in OPR in 10 series releases.
When reading the mini, events are processed at
about 10 Hz.
In principle, almost any work that is done with digis can be
done more simply and quicker with the mini. For this sample job, I have added
the OPR
monitoring module DrcBck to the path to produce some
histograms. These seem to be reasonable. Other modules may be added to
this path to, for example, re-run the fitting or make an event
display. Most of the mini is not currently be kept on disk. If the
collection you want is not on disk, you can retrieve it
with:
> collstagein -include esd "collection name"
The collection used in this example is currently on disk, so
you shouldn't need to stage anything in. So, without any further
ado...
How To Produce the Executable
>newrel -s $BFROOT/work/c/convery/ -t 10.2.3b 10.2.3b
(Use your own scratch area, not mine)
>cd 10.2.3b
>srtpath 10.2.3b Linux2
>addpkg workdir
>gmake workdir.setup
>addpkg Bear
>addpkg
DrcReco V00-06-86-04
>addpkg DrcData V00-06-68-06
>emacs Bear/AppUserBuild.cc
(Add the lines
#include "KalmanTrack/KalFit.hh"
and
add( new KalFit("KalFit","Kal Fit"));
(this must be added after SkimTagSequence(this))
>physboot1
>setenv BDB_SINGLE_CONTEXT n
(I had trouble working in the bridge federation, not sure why.)
>gmake lib
>gmake Bear.bin
How to Run the Job
>cd workdir
>cp ~convery/babar/10.2.3b/workdir/ReadMini.tcl .
>setenv MiniInputCollection \
/groups/Stream1/0001/8200/P10.0.2bV01fb/00018205/cb001/allevents
>setenv MiniHistoFileName recipe.hbook
>../bin/Linux2/BearApp
> source ReadMini.tcl
> ev beg -nev 100
> exit
Then, enjoy your histograms in recipe.hbook
(in the subdirectory drcbck) using PAW or other
browser.
How to Do More
(This section under construction)
Any other reconstruction or monitoring code may be added to the path
in ReadMini.tcl.
Since the executable is Bear, just about any module you would
want to use is already present and just needs to be added to the path. If you
want to
write your own module, it can access this various lists produced
in DIRC reconstruction. For example, you can access the lists of
DrcRecoHit and DrcTrack with:
DrcRecoHitList* aHitList;
getTmpAList( anEvent, aHitList, IfdStrKey("Default") );
DrcTrackList* aTrkList;
getTmpAList( anEvent, aTrkList, IfdStrKey("Default") );
Mark Convery, Updated 13 Dec 01
How to Do Find the Data
To find out which federation contains your favourite run, use Phil Hart's
perl script:
~philiph/perl/col_fed_find.pl
The script will tell you which slave federation the mini has to be
staged to.
Next, set your OO_FD_BOOT variable to that federation.
Now use collstagein to stage the mini.
To get di-muons for run 23220, for example:
~philiph/perl/col_fed_find.pl 23220
> run 23220 is found in phys19 fed in collection
> groups/AllEvents/0002/3200/P10.2.3gV01fb/00023220/cb001/allevents
setenv OO_FD_BOOT
/nfs/objyboot2/objy/databases/production/boot/physics/V1/ana/0119/BaBar.BOOT
collstagein -include esd -wait
/groups/Stream2/0002/3200/P10.2.3gV01fb/00023220/cb001/allevents
collstagein -include esd -wait
/groups/Stream2/0002/3200/P10.2.3gV01fb/00023220/cb001/allevents
Yes, it looks like you still have to call collstagein twice, once to
stage the headers, the second time to stage the "digis".
Then define the environment variables for your job and run your application.
Joe Schwiening, Updated 20 Mar 02
Links:
This page is maintained by
Jochen Schwiening
|
