Linear Collider Forum



Home » Analysis and Reconstruction » Tracking & Vertexing » LCIO::EVENT::TrackerHit quality information
LCIO::EVENT::TrackerHit quality information [message #1916] Wed, 03 March 2010 10:45 Go to next message
rosemann
Messages: 41
Registered: March 2009
Location: hamburg.de
Hi,

in the recent LCIO version there is an additional data member for the Trackerhit; the quality (with accessor/mutator). The layout for the usage was proposed formerly, but seems incomplete to me:

  • bit 0: contains at least one broken channel.
  • bit 1: at least one pulse contributing to the hit is not OK.
  • bit 2: hit at the border of the pad plane.
  • bit 3 - 19: not used for the time being.
  • bit 20 - 30: reserved for private usage.
  • bit 31: not used for the time being


Existing reconstruction code can sometimes be more specific than that. What comes to my mind is:
  • contains a dead pad
  • contains an overflow pulse
  • contains a pulse that has been split
  • contains a pulse with strange shape
  • is at the edge/border


Did I forget something? I would propose that the most usual defects get a bit assigned and only the least part is undefined.


Cheers,
Christoph


When you have eliminated the impossible, whatever remains, however improbable, must be the truth. (Sir A.C. Doyle in Sign of Four)
Re: LCIO::EVENT::TrackerHit quality information [message #1917 is a reply to message #1916] Wed, 03 March 2010 12:17 Go to previous messageGo to next message
killenberg
Messages: 125
Registered: July 2005
Location: CERN
Hello Christoph,

your proposal looks good, but I have some comments:
  • contains a dead pad = bit 0: contains at least one broken channel

  • contains an overflow pulse
  • contains a pulse that has been split
  • contains a pulse with strange shape

These three obsolete bit 1: at least one pulse contributing to the hit is not OK. Or do you have anything else that could be wrong with a pulse?

  • is at the edge/border = bit 2: hit at the border of the pad plane.


I also have some missing items:
  • channel is noisy
  • contains non-dead channel without pulse
  • is multi hit candidate (for instance contains local minimum in charge distribution, or bit is set by topological hit/track finder, which is planned when track candidates are split)

Should we reserve a sequence of bits to store the number of possible hits that contributed?

With this the new proposal looks like:
  • bit 0: contains at least one broken channel (dead)
  • bit 1: contains at least one noisy channel
  • bit 2: hit at the border of the pad plane.
  • bit 3: contains an overflow pulse
  • bit 4: contains a pulse that has been split
  • bit 5: contains a pulse with strange shape
  • bit 6: contains non-dead channel without pulse
  • bit 7: is multi hit candidate
  • bit 8 - 19: not used for the time being.
  • bit 20 - 30: reserved for private usage.
  • bit 31: not used for the time being


Cheers
Martin


Martin Killenberg

CERN
martin.killenberg@cern.ch
Re: LCIO::EVENT::TrackerHit quality information [message #1918 is a reply to message #1917] Thu, 04 March 2010 01:24 Go to previous messageGo to next message
rosemann
Messages: 41
Registered: March 2009
Location: hamburg.de
Hi Martin,

I just wanted to collect the ideas first. I guess the different "qualities" lead to different ways of handling, therefore I wanted to have a finer granularity.

An addition: if we want to use a flag for "multiple hit candidate", then we should also have a flag for "is a split hit" (I guess).


Cheers,
Christoph


When you have eliminated the impossible, whatever remains, however improbable, must be the truth. (Sir A.C. Doyle in Sign of Four)
Re: LCIO::EVENT::TrackerHit quality information [message #1947 is a reply to message #1918] Mon, 19 April 2010 04:07 Go to previous message
rosemann
Messages: 41
Registered: March 2009
Location: hamburg.de
Hi,

another addition, possibly depending on the outcome of the discussion of "how to do alignment":
- flag if a hit has already been aligned

See this thread for the alignment discussion.

Cheers,
Christoph


When you have eliminated the impossible, whatever remains, however improbable, must be the truth. (Sir A.C. Doyle in Sign of Four)
Previous Topic:Alignment for TPC Modules
Next Topic:LCIO::TRACK type keyword definition
Goto Forum:
  

[ PDF ]

Current Time: Fri Feb 21 10:25:50 Pacific Standard Time 2020
.:: Contact :: Home ::.

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