A possible use of Reco-like objects in Bogus
Shahram Rahatlou
By reco-like object we mean a track object defined in the fast simulation similar to Reco tracks. This is important in order to have the same interface for analysis using both the full simulation via reconstruction and the fast simulation.
I have followed some suggestions by Steve Schaffner to make use of already existing classes in traking packages TrkBase and TrkFitter to come out with a TrkRecoTrk object at the end of simulation.
This is the standard track object used to define a BtaCandidate after the reconstruction chain.
Here you can see how the data are passed through the simu-reco-analysis chain.
In order to create a TrkRecoTrk we define a BgsHelixMaker inheriting from TrkBase/TrkFitMaker class with
only one public member function
TrkRecoTrk* makeTrack( const TrkExchangePar& helixPar, const double step )
TrkExchangePar is the standard object to define a set of helix parameters; given a position X and momentum P we can instanciate helixPar calling
TrkExchangePar helixPar = TrkHelixUtils::helixFromMom( X, P, Charge, Field)
Furthermore we have to define a helix reppresentation class, BgsHelixRep, inheriting from TrkFitter/TrkSimpleRep,
used to "contain" the track, within TrkRecoTrk definition.
The constructor for this class is
BgsHelixRep ( const TrkExchangePar& helixPar, double step, TrkRecoTrk* theTrack )
where theTrack is the one owning this reppresentation.
Take a look at UML diagrams for ALL these classes.
These classes already provide useful tools for error matrix calculation and propagation along the track.
Then a possible approach could be that of using all the information available in both SVT and DCH to create a unique charged track.
We could also try different tracks for each detector and combine them later, but it doesn't seem to be very fast as we want.
E-mail: Shahram.Rahatlou@Roma1.infn.it, Tel. +39-6-4991-4236
|
