GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung
Home » PANDA » PandaRoot » Analysis » Invariant Mass Distribution Widening after 4C Fit
Invariant Mass Distribution Widening after 4C Fit [message #18356] Wed, 08 July 2015 09:37 Go to next message
André Zambanini is currently offline  André Zambanini
Messages: 17
Registered: February 2012
Location: FZ Jülich
occasional visitor
 From: *ext.kfa-juelich.de
Hi everyone,

during my investigation of cascade reconstruction with PandaRoot, I found that after the four-constraint fitter my mass resolution got worse. You can see the difference in the plot I attached. Shown are the invariant mass distributions after applying a cut on the probability (>0.01) of the four-constraint fit, once with the pre-fit values (blue) and once with the values after the fit (green).

My decay is p̅p → Ξ̅Ξ → Λ̅π⁺ & ΛK⁻ with Λ → π⁻p and Λ̅ → π⁺p̅. So in my reconstruction I'm dealing with neutral particles (Λ). I asked a bit around here in the institute and it seems, others see a similar effect when having a mixture of neutral and charged candidates as input to the 4C fitter.

I'm running with trunk r27694 but it has been seen from others here with scrut14 as well. More details like a minimal running example will follow as soon as I have concluded my thesis, which is due in the next days. In the meantime: did anyone else see something like this? If so, is there an explanation to that?

Best,
-André

PS: The mass values are taken from the PndRhoTupleQA method qaCand().

index.php?t=getfile&id=8454&private=0
  • Attachment: dm_xi.png
    (Size: 94.07KB, Downloaded 711 times)

[Updated on: Wed, 08 July 2015 09:40]

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18357 is a reply to message #18356] Wed, 08 July 2015 10:03 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant
From: *ip71.fastwebnet.it
What are you constraining exactly? Are you sure all the particles you are constraining have a covariance matrix?

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18358 is a reply to message #18356] Wed, 08 July 2015 10:10 Go to previous messageGo to next message
André Zambanini is currently offline  André Zambanini
Messages: 17
Registered: February 2012
Location: FZ Jülich
occasional visitor
 From: *ext.kfa-juelich.de
I'm constraining the fit probability that I get from curCand->GetFit()->GetProb() but I do that for both distributions. The only difference are the particle's parameters that are modified for the fit.

What the covariance matrix is concerned I don't know. I didn't store this information and I'll probably don't get to it before the 20th.

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18359 is a reply to message #18358] Wed, 08 July 2015 10:18 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant
From: *ip71.fastwebnet.it
Ok, but what is curCand?

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18360 is a reply to message #18356] Wed, 08 July 2015 10:29 Go to previous messageGo to next message
André Zambanini is currently offline  André Zambanini
Messages: 17
Registered: February 2012
Location: FZ Jülich
occasional visitor
 From: *ext.kfa-juelich.de
I combine my reconstructed Ξ+ and Ξ- into the full system:

RhoCandList XibarXiSystem;
XibarXiSystem.Combine(aXiplusFitList, XiminusFitList);
for ( int i_sys = 0; i_sys < XibarXiSystem.GetLength(); i_sys++ ) {
    curCand = XibarXiSystem[i_sys];
    Pnd4CFitter fit4C(curCand, ini);
    fit4C.Fit();
    ...
}

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18365 is a reply to message #18360] Wed, 08 July 2015 11:26 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant
From: *ip71.fastwebnet.it
It would be better to send all the macros you are using, to reproduce the problem.

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18366 is a reply to message #18356] Wed, 08 July 2015 11:33 Go to previous messageGo to next message
André Zambanini is currently offline  André Zambanini
Messages: 17
Registered: February 2012
Location: FZ Jülich
occasional visitor
 From: *ext.kfa-juelich.de
I agree. However, I think my setup is not that easy to get running, so I would make a minimal example out of it. And this I would rather do after handing in my thesis.

My original intention was to ask around first if anybody else saw this in the past. The hunt for the source of this feature I wanted to postpone, otherwise I fear my deadline is endangered.

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18367 is a reply to message #18366] Wed, 08 July 2015 11:38 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant
From: *ip71.fastwebnet.it
Are you using the standard ideal PR the new NewLinks code? The latter was never tested for analysis.

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18368 is a reply to message #18356] Wed, 08 July 2015 11:41 Go to previous messageGo to next message
André Zambanini is currently offline  André Zambanini
Messages: 17
Registered: February 2012
Location: FZ Jülich
occasional visitor
 From: *ext.kfa-juelich.de
The latter, so PndMCIdealTrackFinderNewLinks(). But again, it has been seen by others here in Juelich as well, also with the default ideal pattern recognition.

Report message to a moderator

Re: Invariant Mass Distribution Widening after 4C Fit [message #18372 is a reply to message #18368] Thu, 09 July 2015 15:50 Go to previous message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant
From: *to.infn.it
Which momentum smearing you added to the PndMCIdealTrackFinderNewLinks task? Maybe the covariance matrix is not properly filled there. At least it was never tested.

Report message to a moderator

Previous Topic: Chi2 of kinematic and vertex fitters
Next Topic: entries at unphysical mass after vertex fit
Goto Forum:
  

[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Sun Dec 01 04:11:33 CET 2024

Total time taken to generate the page: 0.00865 seconds
.:: Contact :: Home :: Legal Notice :: Data Protection Policy ::.

Powered by: FUDforum 3.1.3.
Copyright ©2001-2023 FUDforum Bulletin Board Software