Validation of the TrkExtrapolation and TrkFitter packages

Involved Persons

person.gif Tom Atkinson person.gif Anna Franckowiak person.gif Wolfgang Liebig person.gif Andreas Salzburger


Main purpose is the stepwise validation of both, the newly developed track extrapolation package such as the various different track fitters that build (partly) on top of it. The idea is to develop a framework using the most simple setup and make it flexible in a way, such that it can be re-used for varius different setups and higher complexity.

The development and first tests of this validation chain should be done using the most simple setup:

  • StraightLinePropagator
  • KalmanFitter
  • single tracks

Phase 0: The dummy test

Straight Line Tracks created with the InDetMC_Simulation without hit smearing and material interaction should be refitted with the KalmanFitter, $ \chi^2 = 0 $ should be achieved.

For this purpose the TrackStatistics Algorithm inside the InDetFitterValidaion should first have validated that tracks from InDetMC_Simulation are meaningful in respect to number of pixel hits. number of sct hits.

For debugging of eventual problems in the Extrapolation, the SeededPropagationToSurface (from the InDetExtrapolationValidation package) algorithm has to be adopted to be able to steer any possible propagation from one surface to another, by passing the surface type, parameters of both, starting and destination surface such as the starting parameters through jobOptions.

Phase 1: Smear it

Suppose phase 0 is successfully done, the smearing of the hits on the detector surfaces should be turned on and the focus of the validatin should be shifted from an overall picture to the "on layer" values, i.e. printing residuals, pulls and $ \chi^2 = 0 $ probabilities for the detector elements.

For this purpose the DebugUpdator has to be taken instead of the KalmanUpdator and a analyzation macro in root for the produced Ntuple has to be created.

The refitted track statistics have to be plotted and compared against the simulated ones using again the TrackStatistics algorithm in the InDetFitterValidation package.

Phase 2: Scatter

On top of phase 1 and having a "on-layer" picture of the track, particle interactions during both, InDetMC_Simulation and refitting should be turned on and the exercise of phase 1 should be redone, again concentrating on pull distributions . As the InDetMC_Simulation can only reproduce (so far) purely Gaussian material effects, this should be perfectly reflected in the fitter outcome.

--++ Further steps

If this exercise is done, it can be repeated with different magnetic field, fitter setups and finally tried with full simulated data.

Package Location and Algorithm description

Algorithm/Tool LocationSorted ascending Comments Responsible
PropagationToSurface InDetExtrapolationValidation + root macro Anna
SeededPropagationToSurface InDetExtrapolationValidation   Anna
TrackStatistics InDetFitterValidation + root macro Anna
DebugUpdator InDetFitterValidation   Tom
InDetMC_Simulation InDetMC_Simulation   Andi
TrkExtrapolation TrkExtrapolation   Andi
KalmanFitter TrkKalmanFitter   Wolfgang
RefitAlg TrkRefitAlgs   Wolfgang

-- AndreasSalzburger - 01 Aug 2005

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2007-11-07 - TWikiGuest
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback