Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Migrated to Confluence 4.0

...

I am not an expert neither in FRED nor in the different software packages to analyze data. There are people with a lot more expertise than me on these technical issues (Anders, Michael, Leon...). I joined the collaboration just few months ago. Since i am working with both beamtest and full lat data, i spent some time trying to set some simple procedures which allow me to move from one software package to another easily and in a more or less controlled way.  This is what is presented in this web page.

...

An example of how to use it follows:

Wiki Markup_root \ [0\] .x _ _[_LoadLibraries.C_|^LoadLibraries.C]_ _.x  LoadLibraries.C ("BeamtestRelease-v3r0907p2")_

As it will be shown in the next section, one can also include this macro in another macro and execute it from there:

...

/afs/slac.stanford.edu/g/glast/applications/install/@sys/usr/bin/fred

However, I would suggest to run it  with the flag  --log DEBUG, so that a logfile ($HOME/fred.log) is created, and you can try to figerout what happens if . This log file is very useful when  FRED crashes or does not want to run. Besides that logfile, a directory $HOME/.fred with is created and contains other logfiles are created. But those ones did not help me when FRED was not working due to configuration errors...

You can use the simple script myfred (from  csh or tcsh) as an example of how to run the fred version you want  (in debug mode) with the ruby package you want. 

FRED will be acting as a client of Gleam which will be the engine processing the events.
The configuration of the execution of Gleam is performed via the FRED GLAST pannel.

...

  1. A "Gaudi" executable. This is the Gleam executable, Gleam.exe; and its location depends on the software package that is used. BeamtestRelease-v3r0907p2: /nfs/farm/g/glast/u09/builds/rh9_gcc32opt/BeamtestRelease/BeamtestRelease-v3r0907p2/Gleam/v6r18/rh9_gcc32opt/Gleam.exe; EngineeringModel-v6r070329p16: /nfs/farm/g/glast/u09/builds/rh9_gcc32opt/EngineeringModel/EngineeringModel-v6r070329p16/Gleam/v6r8p1/rh9_gcc32opt/Gleam.exeunmigrated-wiki-markup
  1. *A job options ascii file.* This file defines, among other things, the input data and geometry to be used. It is slightly different for different software package versions. Examples follow: [jobOptions_beamtest.txt|^jobOptions_beamtest.txt], [jobOptions_beamtest_montecarlo.txt|^jobOptions_beamtest_mc.txt]\[ , jobOptions_EngineeringModel.txt\|^jobOptions_EngineeringModel.txt\]
  1. A working directory. Any (valid) directory will work

Important Remark

FRED requires to run the script in a BASH shell, therefore the sripts used in the GLAST panel must BASH scripts.

...

  • BeamtestRelease:  ApplicationMgr.ExtSvc  += "EventSelector/EventSelector","EventCnvSvc";
  • EngineeringMode:  ApplicationMgr.ExtSvc  +=  "GlastEventSelector/EventSelector","EventCnvSvc"

 

Now we are at the point in which we have to push the RUN button. If the files you typed in exist and make sense, then you will be able to start displaying events with FRED.

Important Remark

Occasionally one finds events which cannot be properly displayed and SCREW UP the session.  Leon found a problem in AcdReconFiller (2006/07/13) which caused many of those problems (see corresponding discussion in the beam test mailing list). This is fixed in HepRepSvc v0r18p3. Yet it seems that some large events still have problems to be displayed... There are two "dirty" ways of avoiding the problem:

  •  Not reading the mc.root file; this seems to prevent the problem to happen.
  •  Skip those "special" events that cause the problem in a particular run. This can be done if the events are selected by typing the event number in the event number display of FRED (upper part of panel) instead of scanning them with the "play" button.

 

 I think that's all what you need to know to inspect the data with FRED... at least that is all I know about displaying data with FRED. Hope that is useful. Do not hesitate in contacting me if there are things which are not clear or you have questions about this stuff (dpaneque@slac.stanford.edu).