Linear Collider Forum



Home » Analysis and Reconstruction » Tracking & Vertexing » calo TrackState along wrong track direction...?
calo TrackState along wrong track direction...? [message #2351] Fri, 16 January 2015 01:26 Go to next message
jeans
Messages: 22
Registered: November 2012
I have a single 17GeV electron in ILD, which make a nice track with TPC, vtx, sit, etd hits.

I'm interested about it's state at the calorimeter, so I use getTrackState( TrackState::AtCalorimeter ).

This track state occasionally seems to be at the wrong intersection of the helix with the calorimeter, as if the particle goes the wrong way along the trajectory.

Using the TrackState AtLastHit seems OK.

I'm using MarlinReco v01-08

Are there anythings I can test to clarify the situation?
Thanks,
Daniel.
Re: calo TrackState along wrong track direction...? [message #2352 is a reply to message #2351] Mon, 19 January 2015 01:21 Go to previous messageGo to next message
jeans
Messages: 22
Registered: November 2012
Dear all,

I have done some more investigation, and I think I have found the source of the problem.

It seems that the "CalcXingPointWith" method in ILDParallelPlanarMeasLayer.cc does not take account of the requested "mode" (i.e. forwards or backwards along the track) when calculating the intersection of the track with the surface. This deficiency has even been helpfully noted in the code:
"//SJA:FIXME: do we need to consider the mode here",
to which the answer is, in my opinion, "yes"!
(KalDet/v01-12)


the "createTrackStateAtCaloFace" method of MarlinTrkUtils essentially looks for an intersection of the track in each of the 8 planes of the barrel ecal front surface. If it finds one of these, it accepts it (taking the "best" one if there are more than one); if not, it goes on to look at the endcaps.
(MarlinTrk/v01-11)

The track I had problems with entered the ECAL just in the endcap; however a (wrong, backwards) intersection with the barrel had already been found, so it did not look for the correct endcap intersection.



Cheers,
Daniel.

[Updated on: Wed, 24 June 2015 22:18]

Re: calo TrackState along wrong track direction...? [message #2364 is a reply to message #2352] Thu, 25 June 2015 00:14 Go to previous messageGo to next message
gaede
Messages: 232
Registered: January 2004
Location: DESY, Hamburg
Dear Daniel,

thanks a lot for your detailed bug report. We will try and address the problem (for the next ilcsoft release).
Can you point me to an example LCIO file where this rare problem occurs ?

Cheers, Frank.
Re: calo TrackState along wrong track direction...? [message #2365 is a reply to message #2364] Thu, 25 June 2015 18:51 Go to previous message
jeans
Messages: 22
Registered: November 2012
Dear Frank,

I attach an lcio file containing one event.

One track in the MarlinTrkTracks collection (the one with omega = -3.19071e-05 ) has the following TrackState->getReferencePoints:
TrackState::AtLastHit : (1791.98, 319.048, -2273.11)
TrackState::AtCalorimeter: (-1847.35 -219.503 2335.41)

In this case, the AtCalorimeter one has been taken in the wrong direction along the helix.

Regards,
Daniel.

Previous Topic:wrong track direction...? PLEASE IGNORE!
Goto Forum:
  

[ PDF ]

Current Time: Sat Jul 21 20:20:16 Pacific Daylight Time 2018
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.1.
Copyright ©2001-2010 FUDforum Bulletin Board Software