Simulation / Reconstruction Working group

1
Simulation / Reconstruction Working group

• Toby Burnett
• University of Washington
• 11-13 Jan 2000

T
2
Primary goals

• Driven by written requirements
• Clear separation of simulation/reconstruction
• Allow migration to Geant4
• Definition of data formats
• Parameter system
• Integrated Menus and command-line options
• Reconstruction framework

3
Suggested Procedure

• Create a topic list is it complete? Does it
  overlap other working groups? (A proposed one
  follows)
• Topics are usually implemented as packages
• For each item, prepare a document with
• A description of the item
• The current situation
• Requirements for the new version
• Relative priority
• Person who has taken responsibility
• Target date based on priority
• Note that often we can, and should just copy
  someone elses solution, concentrating on what is
  unique to GLAST

4
List of Topics for discussion

• Services
• exception handling
• parameters
• error logging
• persistency
• n-tuple creation
• Detector description geometry
• Non-active
• Active
• Detector/Event/Recon display
• Data formats
• Raw to IOC
• Digis input to recon
• Recon output
• GUI/command line options
• Documentation

• Simulation
• Use GEANT4 (G4)
• Realism Tracking configuration Validation
• Nuclear interactions
• Energy loss fluctuations
• Background models
• Proton direct albedo
• Heavies
• Albedo gammas
• Trigger
• Level 1
• Level 2/3
• Reconstruction
• Framework or Architecture
• Tracker
• Calorimeter
• ACD
• Alignment/calibration

5
Example exception handling

• Description
• Represents need to notify user of an unusual
  situation, with (at least) two levels of
  severity.
• Current practice
• Globals warning and fatal write message to cerr,
  and/or dialog box, aborting in latter case
• Requirements
• Three severity levels info, warning, fatal
• Control over logging
• Easy generation of names
• Discussion
• There is a very complete package from FNAL,
  called Exceptions.
• Priority high
• Responsible person ?
• Target date 1 Feb

6
Short Term Plan

• Until CHEP2000 (Feb 14)
• Finish repackaging, make code available for
  checkout (Toby, Paul)
• Implement dlls, integrated testing (Toby ,
  Sawyer)
• Decide reconstruction architecture (Traudl, Toby)
• Integrate L3 prototype (Saywer, Dan)
• Mar 1
• New Tracker and Calorimeter code checked in and
  verified. (Jose, Arache)
• He, CNO added to flux and gismo interactions
  (Arache, Toby)

7
Basic requirements

• Maintain operational status while
• Improving realism of material and digitization
• Restructuring (packageizing) code
• Introducing an architecture
• Moving all constants to parameter files

8
Whither glastsim?
Geometry
Digitization
Irf
Raw
Simul
Recon
Source
Event
PostRecon (merit)
ntuple?
PSF, Aeff
9
Unsatisfied Requirements (to-do list)

• Simulation
• Nuclear interactions
• Realism
• Energy loss fluctuations
• External geometric description
• Alignment control
• Enable G4 alternative
• Background sources
• Albedo protons (realistic)
• Sky sources
• Reconstruction
• Latest tracking
• Latest CAL
• Define new architecture

• Data
• (Persistent service)
• Raw data
• Event (recon output)
• Parameter handling
• Documentation
• Integrated tests