GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » PANDA » PandaRoot » Tracking » Bug in PndGenfitAdapters
icon4.gif  Bug in PndGenfitAdapters [message #11126] Wed, 27 October 2010 11:33 Go to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant

From: *vpn.unito.it
Exploring the warnings, I have found something very interesting. This is one of the main reasons why I stress that it is important to remove all the warnings, even if they are just warnings!

In particular:

[ 80%] Building CXX object GenfitTools/adapters/CMakeFiles/genfitAdapters.dir/PndGenfitAdapters.o
/home/spataro/test/trunk/GenfitTools/adapters/PndGenfitAdapters.cxx: In function `GFTrackCand* PndTrackCand2GenfitTrackCand(PndTrackCand*)':
/home/spataro/test/trunk/GenfitTools/adapters/PndGenfitAdapters.cxx:36: warning: passing `Double_t' for converting 4 of `void GFTrackCand::addHit(unsigned int, unsigned int, double, unsigned int)'
Linking CXX shared library ../../lib/libgenfitAdapters.so
[ 80%] Built target genfitAdapters  


If you check PndGenfitAdapters.cxx line 36 you find:

retVal->addHit(candHit.GetDetId(),candHit.GetHitId(),0,candHit.GetRho());


In reality, the correct GFTrackCand::addHit function is:

void addHit(unsigned int detId, unsigned int hitId, double rho=0., unsigned int planeId=0);


Then, in the adapter the order criterion "rho" is switched with the detector plane "planeId". We are giving alway the same rho=0, and giving rho as plane id!!

Before committng the correction, I would like to ask to genfit expert if this is correct and in particular if this could affect the kalman global fit performances.

Many thanks.
Re: Bug in PndGenfitAdapters [message #11131 is a reply to message #11126] Wed, 27 October 2010 14:04 Go to previous messageGo to next message
Anonymous Poster From: 129.187.154*
Hi Stefano,

you are right, this is a bug. The call in the adapter should be:

retVal->addHit(candHit.GetDetId(),candHit.GetHitId(),candHit.GetRho(),0);

The planeId parameter has no relevance for the Kalman filter of GENFIT. I do not think that this bug had any bad effects so far, but it should be fixed nevertheless.

Could you change the code, please?

Cheers, Christian
Re: Bug in PndGenfitAdapters [message #11134 is a reply to message #11131] Wed, 27 October 2010 17:42 Go to previous message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant

From: *vpn.unito.it
Hi,
I have committed the fix.
My doubts were not about the plane id, but from the fact that the code was giving always "0" as rho parameter. Is genfit using that number to sort hits or not? In that case, the rho for all the hits was exactly the same -> no sorting (I suppose).
Previous Topic: Charge in FairTrackPar and GeaneTrackRep
Next Topic: Benchmark Tables and plots for tracking performance
Goto Forum:
  


Current Time: Thu Oct 03 22:04:58 CEST 2024

Total time taken to generate the page: 0.00743 seconds