Linear Collider Forum



Home » Software Tools » LCIO » How to merge two LCIO events
How to merge two LCIO events [message #1047] Wed, 05 September 2007 11:17 Go to next message
lima
Messages: 47
Registered: May 2004
Location: DeKalb, IL, USA
Hi,

I would like to know whether there is something out there already for merging hits from two independent LCIO events. I would like
to know of either Java and C++ implementations.

Thanks,
Guilherme
Re: How to merge two LCIO events [message #1048 is a reply to message #1047] Wed, 05 September 2007 11:20 Go to previous messageGo to next message
tonyj
Messages: 138
Registered: January 2004

Something exists for this in the Java implementation -- it is pure LCIO so the resulting merged files can be used by anyone. Jeremy would know the details.

Tony
Re: How to merge two LCIO events [message #1051 is a reply to message #1048] Wed, 05 September 2007 14:33 Go to previous messageGo to next message
NormanGraf
Messages: 80
Registered: January 2004
Hello Guilherme,
As Tony pointed out, this functionality does exist in the lcio distribution. For details see:
http://confluence.slac.stanford.edu/display/ilc/LCIO+Command +Line+Tool#LCIOCommandLineTool-merge

Norman
Re: How to merge two LCIO events [message #1054 is a reply to message #1047] Thu, 06 September 2007 08:04 Go to previous messageGo to next message
gaede
Messages: 233
Registered: January 2004
Location: DESY, Hamburg
Hi Guilherme.

there is a new package for Marlin called Overlay that provides this functionality for Marlin. You can specifiy the number of events that are to overlayed per 'physics' event either fixed or from a poisson distribution.
You can download the package at:

http://www-zeuthen.desy.de/lc-cgi-bin/cvsweb.cgi/Overlay/?cv sroot=marlin


Frank.
Re: How to merge two LCIO events [message #1070 is a reply to message #1047] Thu, 06 September 2007 12:42 Go to previous messageGo to next message
killenberg
Messages: 125
Registered: July 2005
Location: CERN
Hello,

I would like to point out that merging events is not that simple with a TPC. This is due to the fact that the third coordinate which is measured is time. So one has to take into account the drift velocity and (temporal) bunch spacing to calculate the correct time where the TPC sees the hit. One also has to keep in mind that the TPC only takes one large picture of the whole bunch train and one has to disentangle later which track belongs to which event.

To calculate the occupancy (which I suppose is the reason why you want to overlay events) one further has to take into account the readout frequency and the diffusion in the gas to calculate the correct number of occupied voxels.

In my opinion it does not make much sense to simply merge SimTrackerHits from different events for a TPC, since one can only obtain reasonable tracks if one bypasses major parts of the digitisation and reconstruction and more or less directly fits the track to the SimTrackerHits (maybe with a little smearing). But I doubt that such a simplified ways is suited for an occupancy study.

Track reconstruction efficiency etc. can only be tested if the tracks have the correct 'position' in time, anyway.

The MarlinTPC group has just started working on a pileup processor for the TPC to get a realistic digitisation. Is there a similar solution available in Java?

Greetings

Martin


Martin Killenberg

CERN
martin.killenberg@cern.ch
Re: How to merge two LCIO events [message #1072 is a reply to message #1047] Thu, 06 September 2007 13:53 Go to previous messageGo to next message
NormanGraf
Messages: 80
Registered: January 2004
Hello All,
I don't know why, but I was assuming that Guilherme was asking about merging the output from the full detector simulations, i.e. SimCalorimeterHits and SimTrackerHits. For this, the lcio merge package correctly accounts for time offsets, so that further digitization can do the right thing. I believe it is impossible to correctly merge hits further down the line (e.g. TrackerHit and CalorimeterHit). One can come up with technical ways of doing so, but they all have inherent problems. Martin has pointed out some of them for the TPC, but there are similar problems for other subdetectors. I would strongly encourage people to restrict themselves to overlaying the raw output and doing the digitization correctly.
Norman

[Updated on: Thu, 06 September 2007 14:05]

Re: How to merge two LCIO events [message #1134 is a reply to message #1047] Sun, 23 September 2007 21:49 Go to previous message
antonio.bulgheroni
Messages: 66
Registered: January 2007
Location: INFN - Roma3

Dear all,
I was looking for a smart way of merging the content of LCEvent coming from different files, when I found this post.

I had a look and install the Overlay Marlin package, but I don't think it fulfills may needs.

What I would like to do is in the following:


  1. During a test beam effort, it is very luckily that the tracking hodoscope (our EUDET-JRA1 telescope) is saving its output into a LCIO file, while the Device Under Test is steered by its own DAQ and it is saving its output into another file.
  2. At a certain point we need to "overlay" the two files (say that are both LCIO) but in a very precise way keeping the run number and the event number perfectly synchronized.
  3. Moreover, the merged event should contain all the collections available in the two input files.


Is it possible to do something like that? I have the impression that Overlay is not the right tool to do it, but can it be tailored? Should we write something from scratch?

Thanks for your help !!!

Antonio
Previous Topic:patch release v01-08-04
Next Topic:lcio for whizard
Goto Forum:
  

[ PDF ]

Current Time: Mon Oct 21 17:45:04 Pacific Daylight Time 2019
.:: Contact :: Home ::.

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