Linear Collider Forum



Home » Analysis and Reconstruction » EUDET Telescope » A Marlin processor for pedestal calculation
Re: A Marlin processor for pedestal calculation [message #701 is a reply to message #693] Fri, 09 February 2007 02:46 Go to previous message
gaede
Messages: 233
Registered: January 2004
Location: DESY, Hamburg
Hi Antonio,

I am glad that the 'rewind feature' works for the pedestal calibration, at least for the best part of it. As pointed out this is somewhat experimental software that I quickly implemented. For a release I would like to iterate on the design and understand the implications and use cases better. For example I think an additional callback 'calibrateEvent(LCEvent* evt)' would make the usage a lot simpler and clearer.

Putting the rewind feature also to the end() method would cause potentially severe problems with resource allocation. The modular design of Marlin provides the init() and end() callbacks to the processors for resource allocation and it is guaranteed that both methods are called exactly once before and after the processing of events respectively.
Also there is currently no way to tell how many events are going to be in the input data stream that is going to be processed (other then reading all event headers in the stream).

I think both things are not really needed for a calibration. One should have a well defined criterion to end the calibration loop and rewind, e.g. the first N events or until all errors of constants are below a certain level.

However if you really want to run the calibration on all events
you can easily do this right now by simply specifying the input stream twice and toggling the calibration flag if the first event is presented for the second time.

If you want to make your calibration persistent I would suggest to use LCCD the conditions data toolkit. With this you can for example store the calibration in a plain LCIO file and then easily access the calibration in a Marlin job that processes the events.
LCCD has various options on how the conditions (calibration) data are stored/read: database, plain LCIO file, LCIO file with DB dump or in the Data file directly. Please check the documentation and examples.

Cheers, Frank.



 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic:Important steps toward the EUDET telescope analysis
Next Topic:ilcinstall: configuration file for Eutelescope
Goto Forum:
  


Current Time: Sat Sep 22 12:42:46 Pacific Daylight Time 2018
.:: Contact :: Home ::.

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