GSI Forum: Bugs, Fixes, Releases » RIEMANN & LHETRACK fight

Home » PANDA » PandaRoot » Bugs, Fixes, Releases » RIEMANN & LHETRACK fight

Show: Today's Messages :: Polls :: Message Navigator

RIEMANN & LHETRACK fight [message #6358]

Thu, 10 April 2008 19:48

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *physik.uni-giessen.de

Hello dear code developer.
I would like that you take a look into my problem with riemann+lhetrack, because I could not find a clue (up to now).
If you download the last svn release, you can try exactly my macros.
Go to macro/lhetrack, and launch

run_sim_tpcmvd.C
run_track_tpcmvd.C
plot_pT.C

You will have a nice peak, that's fine. I have tried also at GSI on lxg0442 and it works.
Now, please edit run_track_tpcmvd.C, uncomment out the riemann part:

PndTpcRiemannTrackingTask* tpcSPR = new PndTpcRiemannTrackingTask();
tpcSPR->SetTrkFinderParameters(2.,// proxcut
0.02, // proxcut on rieman sphere
2.E-3, // planecut
4.0, // szcut
4); // minnumhits for fit
tpcSPR->SetPersistence();
fRun->AddTask(tpcSPR);

and rerun run_track_tpcmvd.C. It will work, I have tried also at GSI on lxg0442 .

Now run plot_pT.C: it will crash!!!
But lhetrack code does not ask for riemann data object. You can set lhetrack working only with TpcPoints (SetTpcMode(1), SetMvdMode(0)) and you will have the same crash, not while analysing the event but when accessing to the files via macro. This is the error at the t->GetEntriesFast() point:

Toggle Spoiler

If you try something via TBrowser, such as:

cbmsim.Draw("PndTpcLheTrack.fP.Perp()","PndTpcLheTrack.fP.Perp()<2")

it will work.

What is going wrong? Could you please try to reproduce this error?
I want to stress that this problem appeared to me the first time with the old lhetrack code, before the last changes. So it is not my fault Smile