Difference: LHCbPhysicsEventModelTaskForce (29 vs. 30)

Revision 302006-11-17 - unknown

Line: 1 to 1
 
META TOPICPARENT name="LHCbComputing"
Changed:
<
<

LHCb Physics Event Model Task Force Page

>
>

LHCb Physics Event Model Task Force Page

 

Members

Line: 34 to 35
 

Scope

Here is a first brain dump of the "scope" by PK
Changed:
<
<
  • The essential ingredients are of course the Particle and Vertex classes, as well as their MC-true counterparts (that affect all software, not only DaVinci), but it would be nice if we could also look at the other "physics" classes like the new FlavourTag class for instance.
  • The MC part is more urgent as we need to start production early 2006. For the pure DaVinci stuff, more iterations are allowed.
>
>
  • The essential ingredients are of course the Particle and Vertex classes, as well as their MC-true counterparts (that affect all software, not only DaVinci), but it would be nice if we could also look at the other "physics" classes like the new FlavourTag class for instance.
  • The MC part is more urgent as we need to start production early 2006. For the pure DaVinci stuff, more iterations are allowed.
 
  • Now that the new Track class is available a change of the ProtoParticle is mandatory. We'll have to revise how we link to PID and how of much of it we store in the ProtoParticle itself (or if we can get rid of it).
    • (MC) Perhaps we should also define the DST content at the same time as re-defining the ProtoParticle.
  • How do we relate a Particle to the PV that has been used to select it (if any)?
  • One question we'll have to have in mind is how these objects get persistified and what is the minimal amount of data one needs to make a physics analysis (please don't answer: "the raw buffer").
  • Changes in the event classes cannot be decoupled from changes in the interfaces for the basic tools. For instance the vertex fitters interface depends on how the Particle and the Vertex are connected. We might hence also have to have a look at the most basic tools: the VertexFitter, PhysDesktop, ParticleMaker, FilterCriterion...
Changed:
<
<
  • This will bring to the question of how the core DaVinci tools and LoKi should be related... I don't know if we'll have time to address that.
>
>
  • This will bring to the question of how the core DaVinci tools and LoKi should be related... I don't know if we'll have time to address that.
 

Additional comments

  • (VB) Do we want to have the same physics event model for "off-line" and HLT? (personally i hope that the answer is "yes")
Line: 307 to 308
  One should also have a look at the basic tools. We have to make sure that the fucntionality removed from the event model is provided by some helper tool. Here is a list of requirements to the tools using the event model.
  • Vertex fitters that modify the input particles sould have a method taht returns a clone of these particles, so that the they can be stored in the TES.
    • The PhysDesktop already does that automatically
Changed:
<
<
  • Once the review is over, Patrick and Vanya have to sit together to review the core DaVinci packages and LoKi. A merger of LoKi::Algo and DVAlgorithm is desirable.
>
>
  • Once the review is over, Patrick and Vanya have to sit together to review the core DaVinci packages and LoKi. A merger of LoKi::Algo and DVAlgorithm is desirable.
 
  • Comment about photons: The PhotonTool is essentially an extrapolator for photons. It should be used inside the extrapolator, not the user code.
  • We would like to be able to write out a "nano-DST" containing only the selected particles but with enough information to redo the fits. For instance the B, the full decay tree and everything needed to make it.
  • We would like to be able to save the options that have been modified when creating a DST in the DST itself. This is the only way to ensure that one does not loose the conditions in which it has been created.
Line: 322 to 323
 
  • 08/10: Updated Particle, ProtoParticle and Vertex classes following the discussion on the August 9 meeting.
  • 08/16: Starting to modify MC truth.
  • 08/29: More about MC truth. Added a LHCbTypedefs file with some useful typdefs. They are used throughout the event model.
Changed:
<
<
  • 09/09: Started implementation of Minimal DaVinci. DaVinciKernel, DaVinciFilter, VertexFitter (minimal version) and DaVinciTools are done. Missing are the more tricky ParticleMaker and GlobalReco
>
>
  • 09/09: Started implementation of Minimal DaVinci. DaVinciKernel, DaVinciFilter, VertexFitter (minimal version) and DaVinciTools are done. Missing are the more tricky ParticleMaker and GlobalReco
 

Next steps:

  • Implement MC truth
Changed:
<
<
  • Compile minimal DaVinci: ChargedProtoPAlg, CombinedParticleMaker, DVAlgorithm, PhysDesktop, Extrapolators, introduce new Vertex fitters.
  • Test minimal DaVinci: Build some J/psi.
>
>
  • Compile minimal DaVinci: ChargedProtoPAlg, CombinedParticleMaker, DVAlgorithm, PhysDesktop, Extrapolators, introduce new Vertex fitters.
  • Test minimal DaVinci: Build some J/psi.
 

Trash (of the TF)

Line: 346 to 347
 
Changed:
<
<
>
>
 
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback