DaVinci Migration to DC06 page

Timescales

Code needed for DC06 stripping

The DC06 stripping has started. All code necessary for stripping is there. Some pieces of code needed for analysis are not yet done.

To do:

• Test all options-based pre-selections
• Continue to review all code.
  • Remove obsolete stuff.
  • Avoid duplications of code. Always! No exceptions!
• More details in the various packages.

Desired features non essential for DC06 stripping.

• Split LoKi in pieces: Done.
  • Merge DVAlgorithm and LoKi::Algo as much as possible: Done.
  • Remove code duplications. In particular LoKi::Cuts and DaVinciFilter: Being done by HLT group.
• Adapt to python options: There is a hope.
• Unify string parsing. Might not be needed anymore once LoKi::Cuts and DaVinciFilter are merged.

Meetings

• The forum to discuss DaVinci code is T-Rec.
• Long ago: LHCb week, Tuesday March 7 2006. http://indico.cern.ch/conferenceDisplay.py?confId=1196. Phone: +41 22 767 7000 (77000) (old system). VRVS: Sun virtual room.

Status history

The doxygen of the latest version released of PHYS is available at http://lhcb-release-area.web.cern.ch/LHCb-release-area/PHYS/doc/html/index.html

• June 2007: Stripping OK, but some bits for analysis not yet there.

• 21/07/2006 Essential code runs.
  • Test stripping jobs select random events from rDST and produce event tag collections
  • New ParticleMakers
  • DaVinciAssociators
  • VertexFitters
  • CommonParticles tested for long-lived particles and J/Psi

• 29/05/2007: Essential core code compiles. DaVinci initializes and finalizes properly.
  • Working versions of MCParticleMaker, OfflineVertexFitter and BlindVertexFitter running and roughly tested with Gauss events.
  • Includes new LoKi packages.
  • Options-based DC06 pre-selection packages tuned on DC04 imported but not tested.

See doxygen to-do list and interface status.

Packages status

Warning: DC04 developments sit on a branch CVS branch DC04b. DC06 developments are in head.

Phys/DaVinciKernel

In charge: DaVinci Librarian

Purpose: Contains all interfaces andDVAlgorithm.

Status: OK

Phys/DaVinciTransporter

In charge: Patrick K.

Purpose: Transporter (extrapolator) tools for particles. Used to contain a lot of duplications of track extrapolators, but for Particles. Now that the track extrapolators extrapolate States one can translate any Particle to a State and back.

Status: Still not optimal. Needs scrutiny.

Do not duplicate extrapolators code!

Phys/ParticleMaker

In charge: Global Particle ID Task Force

Purpose: Contains all tools making Particles from ProtoParticles.

Status: OK

Phys/VertexFit & Phys/VertexFitChecker

In charge: Yuehong Xie

Purpose: Vertex fitters and re-fitters.

Status: Mass constrained vertex fitting still under discussion.

Phys/DaVinciFilter

In charge: DaVinci librarian

Purpose: All FilterCriteria tools, the ParticleFilter tool and the FilterDesktop algorithm.

Status: OK

To do: Replace all filter by HybridFilterCriterion to be used in the HLT as well.

Phys/DaVinciTools

In charge: DaVinci Librarian

Purpose: All "other" tools and algorithms. In particular the MakeResonances algorithm.

Status: OK

To do:

• Replace MakeResonances by CombineParticles using Hybrid filter. Then slowly translate options and retire MakeResonances.

Phys/CommonParticles

In charge: Olivier Deschamps and DaVinci Librarian

Purpose: Package containing code to produce standard particles. Many options but also C++ for pi0 making.

Status: OK

Phys/FlavourTagging & Phys/FlavourTaggingChecker

In charge: Marco Musy

Purpose: Flavour tagging.

Status: OK.

Phys/DaVinciAssociators

In charge: Philippe Charpentier

Purpose: (Proto)Particle to MCParticle association code.

Status: works more or less. Some strange features to be understood.

To do: rewrite.

Phys/DaVinciMCTools

In charge: Stefania Vecchi for the MCParticleMaker. Vladimir Gligorov for BackgroundCategory. And for the rest DaVinci librarian

Purpose: All low-level tools and algorithm that need access to MC truth.

Status: Seems OK.

Phys/DaVinciEff

Orphan package

Purpose: Efficiency algorithms.

Status (20/04): Nothing done.

To do:

• Adapt to new event model.

Phys/Tampering

In charge: Mariusz Witek ?

Purpose: Buffer tampering algorithms.

Status: Nothing done. Wait for trigger to be there.

Phys/TriggerSource

In charge: Patrick Koppenburg

Purpose: TisTosTobbing without Tampering algorithms. An alternative to Tampering.

Status: Nothing done. Wait for trigger to be there.

Tools/Stripping

In charge: DaVinci Librarian

Purpose: Stripping algorithm and python scripts.

Status: DC06-compatible v5r1. C++ compiles. Python produces correct stripping job options.

To do:

• Python script needs SERIOUS clean-up.

Links

• LHCb Event Model Review
• Physics Event Model Task Force
• Global Particle ID Task Force
• Flavour Tagging WG

-- PatrickKoppenburg - 07 Feb 2006 -- PatrickKoppenburg - 13 Jun 2007