GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » PANDA » PandaRoot » Tracking » Test of kinematic/vertex fitters
Test of kinematic/vertex fitters [message #15432] Wed, 04 September 2013 16:27 Go to next message
Klaus Götzen is currently offline  Klaus Götzen
Messages: 293
Registered: June 2006
Location: GSI
first-grade participant
From: *gsi.de
Dear tracking experts,


I have a question related to tracking, and I'm sure that somebody of you can give me some suggestion.

What I want to do is to check whether the vertex/kinematic fitters in Rho are working properly. Therefore I'd like to feed them with tracks, which are smeared based on the true tracks from MC and have a reasonable covariance matrix.

I already tried it with a 7x7 (4-vector/position) covariance, only setting the diagonal elements to some values and smearing the 7 components of the candidates accordingly by this diagonal matrix. Unfortunately this didn't work (the probability distributions didn't look as expected, even worse than for full MC tracks). This has perhaps to do with missing correlations (=covariances) between the kinematic components, which implicitely are used by the fitter, e.g. helix shapes etc.

Therefore finally my quesion to you: Does somebody of you have an idea, how I can create reasonably smeared toy MC tracks together with a corresponding 'artificial' covariance matrix to reach my goal? Maybe starting from a 5x5 helix (diagonal?) covariance matrix and smeared tracks in the 5 parameters helix domain, which then is converted to the 7-parameter domain?


Best regards and thanks for any feedback,
Klaus
Re: Test of kinematic/vertex fitters [message #15434 is a reply to message #15432] Wed, 04 September 2013 17:46 Go to previous message
Gianluigi Boca is currently offline  Gianluigi Boca
Messages: 177
Registered: March 2004
first-grade participant
From: *pv.infn.it
dear Klaus,
a concise answer is impossible in the Forum. If you like we can talk on the phone or next week at the meeting
Tschuess!

Gianluigi

Klaus Goetzen wrote on Wed, 04 September 2013 16:27

Dear tracking experts,


I have a question related to tracking, and I'm sure that somebody of you can give me some suggestion.

What I want to do is to check whether the vertex/kinematic fitters in Rho are working properly. Therefore I'd like to feed them with tracks, which are smeared based on the true tracks from MC and have a reasonable covariance matrix.

I already tried it with a 7x7 (4-vector/position) covariance, only setting the diagonal elements to some values and smearing the 7 components of the candidates accordingly by this diagonal matrix. Unfortunately this didn't work (the probability distributions didn't look as expected, even worse than for full MC tracks). This has perhaps to do with missing correlations (=covariances) between the kinematic components, which implicitely are used by the fitter, e.g. helix shapes etc.

Therefore finally my quesion to you: Does somebody of you have an idea, how I can create reasonably smeared toy MC tracks together with a corresponding 'artificial' covariance matrix to reach my goal? Maybe starting from a 5x5 helix (diagonal?) covariance matrix and smeared tracks in the 5 parameters helix domain, which then is converted to the 7-parameter domain?


Best regards and thanks for any feedback,
Klaus

Previous Topic: Start and stop values from PndRecoKalmanTask
Next Topic: [FIXED] Symbol lookup error :: PndMvdRiemannTrackFinderTask
Goto Forum:
  


Current Time: Thu Mar 28 20:38:35 CET 2024

Total time taken to generate the page: 0.00616 seconds