GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » PANDA » PandaRoot » Bugs, Fixes, Releases » analysis for TOF studies, fullsim crashes
analysis for TOF studies, fullsim crashes [message #9183] Tue, 11 August 2009 16:44 Go to next message
Bernhard Roth is currently offline  Bernhard Roth
Messages: 8
Registered: February 2009
Location: Bochum
occasional visitor
From: *ep1.ruhr-uni-bochum.de
Hi,

I am doing TOF studies and also want to compare the results of my analysis between our BaBar-like software and PANDAroot. Therefore I'm analyzing the reaction
pbar p -> Phi Phi Pi0; Phi -> K+ K-;
all PHSP and at 2GeV beam momentum.
As background I'm considering Pi+ Pi- K+ K- Pi0
All with full simulation.
In PANDAroot I want to use the scripts from the PandaRootTutTorino09 which is tutorials/analysis/go_stt_evtgen.sh.
And I'm using the latest stable version from 10.8.2009, running on a GSI 32bit machine.
As first step I had to change run_reco_sttcombi.C:
Toggle Spoiler


Then it crashes whether I run over 10 events or more at different points.
On more events it crashes in run_kalman_stt.C with the following error messages in 4-kalman.log:

Toggle Spoiler


In case of running only over 10 events it finishes run_kalman_stt.C successfully, but crashes in the next step, in makeTCands_stt.C. The error message in 5-microwriter.log is here:

Toggle Spoiler


Did I use the correct macro with correct input?
Can some expert help me in this context?

Cheers,
Bernhard

P.S.
To compare the results between PANDAroot and the BaBar-like software it would be great to have the same data-sets available.
For signal (ppbarsystem -> Phi Phi Pi0, see above) I have now 100k events and for background Pi+ Pi- K+ K- Pi0 10M events for the BaBar-like software available. Is there an opportunity to order such an amount of data for doing analysis with PANDAroot? (all fullsim)
Re: analysis for TOF studies, fullsim crashes [message #9184 is a reply to message #9183] Tue, 11 August 2009 16:56 Go to previous messageGo to next message
Jens Sören Lange is currently offline  Jens Sören Lange
Messages: 193
Registered: June 2005
first-grade participant
From: *physik.uni-giessen.de
Hi Bernhard, a quick idea:

#34 0x40826044 in G__getitem (item=0xbfd46640 "run_kalman_stt(100)") at cint/cint/src/expr.cxx:1896
#35 0x4082387b in G__getexpr (expression=0x82f39e8 "run_kalman_stt(100)") at cint/cint/src/expr.cxx:1470

can it be that this crash happens because you try to run the Kalman over 100 events but you only have 10 events at the input?
P.S. can you maybe post your EvtGen decay files, for signal and background (PHSP should be simple, but just for checking what final state particles with what momenta you have which enter the Kalman ...)

Soeren
Re: analysis for TOF studies, fullsim crashes [message #9185 is a reply to message #9184] Tue, 11 August 2009 17:23 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant

From: *56-82-r.retail.telecomitalia.it
Hi,
about the first kind of errors, the structure of lhetrack was strongly changed with respect to the Torino meeting (therefore from the tutorial).
This means that the task name is a bit different (your change in reco), but I cannot see your crash in kalman macro without full access to the code. The only thing that I can suggest is to comment out the pidmaker in the reco macro (this could give some problems after with the new kalman task).

About the makeTCand, the code in rho was not updated according to the new structure, this is simply the reason why it is not working. I know Klaus was working to update it, but then he left for vacation.

If you want to reproduce your studies, I would suggest to use the svn release for the tutorial, exactly that revision numer (which was before the change).
Re: analysis for TOF studies, fullsim crashes [message #9190 is a reply to message #9183] Wed, 12 August 2009 10:48 Go to previous messageGo to next message
Bernhard Roth is currently offline  Bernhard Roth
Messages: 8
Registered: February 2009
Location: Bochum
occasional visitor
From: *ep1.ruhr-uni-bochum.de
Hi Soeren,

thank you for your fast response!

With EvtGen I have generated 1000 events. They are all in the output.evt file listed and the events looks quit fine.
So this shouldn't be a problem.
My DEC file is:

Particle pbarpSystem 5. 10.

Decay pbarpSystem
  1.0 phi phi pi0       PHSP;
Enddecay

Decay phi
   1.0 K+ K-            VSS;
Enddecay

End

The DEC file is located in
/d/panda02/broth/proot/stable/workdir/dec_files

and the output.evt is
/d/panda02/broth/proot/stable/workdir/gen_files/PhiPhiPi0_2GeV_1M.evt
for running over it I copy and rename it (output.evt) to the tutorial/analysis/ directory, of course

Bernhard
Re: analysis for TOF studies, fullsim crashes [message #9193 is a reply to message #9190] Wed, 12 August 2009 12:41 Go to previous messageGo to next message
Bernhard Roth is currently offline  Bernhard Roth
Messages: 8
Registered: February 2009
Location: Bochum
occasional visitor
From: *ep1.ruhr-uni-bochum.de
Hi!

For reproducing the tutorial I need to know the exact revision of it.
There is written:
Quote:


Whenever you want to update the software to a particular revision (e.g. 5813), execute ...


Therefore I tried revision 5813, but got the same errors.

Does anybody know the revision of the tutorial (PandaRootTutTorino09)?

Bernhard
Re: analysis for TOF studies, fullsim crashes [message #9194 is a reply to message #9193] Wed, 12 August 2009 13:54 Go to previous messageGo to next message
Johan Messchendorp is currently offline  Johan Messchendorp
Messages: 693
Registered: April 2007
Location: University of Groningen
first-grade participant

From: *xs4all.nl
Dear Bernhard,

As far as I know this (5813) was the revision number used for the tutorial.

For the large data production, you can contact Paul Buehler to start jobs on the GRID.
Alternatively, you can also do it yourself. For that contact Dan Protopopescu to get an howto an obtaining an account and the relevant information....

Johan.
Re: analysis for TOF studies, fullsim crashes [message #9195 is a reply to message #9193] Wed, 12 August 2009 14:09 Go to previous messageGo to next message
Bernhard Roth is currently offline  Bernhard Roth
Messages: 8
Registered: February 2009
Location: Bochum
occasional visitor
From: *ep1.ruhr-uni-bochum.de
Hi,

I made now an all new copy of revision 5813 and now fullsim runs trough and finishes successfully!
It seems that PndChargedCandidates are empty but before facing me with that I have to run on more events for doing analysis with them.

Cheers,
Bernhard
Re: analysis for TOF studies, fullsim crashes [message #9199 is a reply to message #9183] Wed, 12 August 2009 16:22 Go to previous message
Bernhard Roth is currently offline  Bernhard Roth
Messages: 8
Registered: February 2009
Location: Bochum
occasional visitor
From: *ep1.ruhr-uni-bochum.de
Hi,

I'm now in revision 5813 and still running the macros from PandaRootTutTorino09 for fullsim of decay
p pbar -> phi phi pi0; phi -> K+ K- at 2 GeV

It finishes successfully on 10 and 100 events, but crashes on 1000 events in run_kalman_stt.C.
The error message in 4-kalman.log is:

Toggle Spoiler


But even if it runs fine, the tuples of PndChargedCandidates are all empty (in cases of running over 10 or 100 events).

Bernhard
Previous Topic: PID package
Next Topic: workaround for PndMicroWriter.cxx
Goto Forum:
  


Current Time: Fri Mar 29 07:44:00 CET 2024

Total time taken to generate the page: 0.00993 seconds