hi blasters,
here is a summary of what i got out of our wednesday discussion about
the class structure of the reconstruction software. it is just skeleton
of some of the revelent classes for the second half of the
reconstruction, with lots missing. for example, how to add other
detector information into the reconstruction is still an open question.
The classes are indented according to inheritance, with descriptions
and relevant members and methods on the right. we found out that tracks
and links have come to meant he same thing, so i propose that links be
anything that gets to the track fitter, while tracks are the final
accepted prongs of an event.
TBLEvent Event header info, written to DST
TBLEvStart,Stop,etc Coda events written to DST
TBLEvEpics Epics events written to DST
TBLEvScaler Scaler events written to DST
TBLEvPhysics Physics events, written to DST
- fNTracks
- TBLTrack fTrack[3] (only physical tracks)
- fPID[3] (corresponding id's)
- fChiSQ (for whole event)
TBLRecon Event reconstruction, track selection, final fit
- TBLTrack fLink[n] (all links w/decent
fastfit)
TBLAnalyze Read from DST and analyzed
TBLTrack WC,SC,Recoil info for one track/link
- TBLWc1Segment fData[3]
- p[5]={p,th,phi,z,d}, charge
TBLFit Info for fitting track parameters
- ChiSq(), fChiSq
- TBLWc1Segment fSim[3]
- fit4(), fit3()
TBLNewt Specific implementations
TBLSmart "
TBLMinuit "
TBLWc1Segment Wire chamber classes
TBLWc1Stub, Hit, ... "
does this look ok? any suggestions?
also, Tim, can you change the permission of cvsroot/BlastLib2 so we
can check it out?
--chris
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST