210 likes | 307 Views
Mokka, main guidelines and future. P. Mora de Freitas Laboratoire Leprince-Ringuet Ecole polytechnique - France Linear collider Workshop 2004, Paris. The first main guideline:. Let’s compare detector designs in a common framework ! It means, to simulate different detector models with :
E N D
Mokka, main guidelines andfuture P. Mora de Freitas Laboratoire Leprince-Ringuet Ecole polytechnique - France Linear collider Workshop 2004, Paris
The first main guideline: Let’s compare detector designs in a common framework ! • It means, to simulate different detector models with : • The same physics • The same simulation framework (Geant4) • The same input data files • The same output data format (to be able to apply almost the same reconstruction and analyze programs) P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
The second main guideline: Keep it simple, stand alone as possible and available for all ! • It means, • to just rely on the Geant4 framework functionalities and standard building tools (gmake) • To write out just ASCII files • To provide an open wide access via the Mokka WEB page P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
The third main guideline: Keep trace ! • It means, • Tagged releases (almost monthly) • Release notes for each release • Just one reference geometry database, indeed local copies can be useful while developing new detector models • Output data refers to a known detector model in the reference geometry database • Run log file (detector model, geant4 and Mokka release tag ids, line command parameters, etc) P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Mokka first architecture and relationships P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
The geometry database P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
The geometry database A detector model sample: “D08” Ingredients Model Sub-detectors P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Impact of the gas in HCAL Electrons Pions P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Physics validation: Calice Y Hcal Prototype (for the moment just aTesla Hcal module build as a 1m3cube ) Ecal Prototype X Beam
What we learned I:Geometry have to be shared Generator Need to share the same geometry !!! Detailed simulation Reconstruction Analysis P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Analysis (user code) Simulation (Mokka) Reconstruction (user code) CGA API Geometry drivers Geometry Database C.G.A: the Common Geometry Access API (F77, C++,C and Java) Run Time • Relies on Geant4 geometry layer • Implements some reconstruction utilities. (By G. Musat) P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
What we learned II:Data format have to be shared Generator Need a common High level data format Detailed simulation Reconstruction Analysis P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
LCIO Persistency Framework Generator Analysis Recon- struction Simulation Motivation Java, C++, Fortran Java, C++, Fortran Geant3, Geant4 Java, C++, Fortran (Slide by Tony Johnson) geometry
JAS3 + LCIO • JAS3 Plugin exists for viewing and analyzing LCIO files • Allows to directly open and browse any LCIO • Works with any LCIO file • no requirement for prior knowledge of what is being read. • Recently updated to support LCIO 1.0 • (in fact 0.8 or later) (Slide by Tony Johnson)
What we learned III:New users ask for new functionalities and use cases ! • Big thanks to Frank Gaede (DESY) for a set of new user facilities for Mokka: • steering files • init commands • environment variables • Plugin interface for Geant4 user actions, mainly for “on line” analyzes P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
But what we really learned:We have to work together ! • Big improvement since the LCIO interface • The LCWS04 Mokka release includes very good contributions also from • Jeremy McCormick (NICADD) • Frank Gaede (DESY) P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Future • We have to improve working together: • by an informal developers committee and its organization around the CVS repository management, code standards, Geant4-LC users interface, etc. • to bring adequate level of detector descriptions to this common simulation tool. Job for detector specialists P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Mokka’s kernel framework Geant4 Mokka Kernel Developer P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Who is who for the next model improvements in Mokka? • Ecal: • W/Si calorimeter and its prototype: detailed model by Ecole polytechnique • Other options: … • Hcal: • Digital and its prototype: • RPC: generic model by Ecole polytechnique, detailed model by … • GEM: detailed model by Nicadd (Jeremy McCormick) • Analog (scintillator, etc): generic model by Ecole polytechnique, detailed model by … P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Who is who for the next model improvements in Mokka? • Inner tracking devices: • Generic model by Ecole polytechnique (directly translated from Brahms), detailed model by … • The SET: LPNHE – Paris + … • Forward detectors (masks, LAT, …): • Generic model by Ecole polytechnique (directly translated from Brahms), detailed model by … • Muon chamber: … • Beam line: … P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris
Conclusions: 1) “Improve framework” but “Keep things simple”. 2) “Let’s compare detector designs in a common framework”, to do this “We need adequate level of detector descriptions”. Let’s work together !!! P. Mora de Freitas – L.L.R. Ecole polytechnique, France – LCWS04 Paris