Skip to content
Snippets Groups Projects

Test for DecayTreeFitterAlg

Merged Test for DecayTreeFitterAlg
pkoppenb-DecayTreeFitterAlg into master
Merged Patrick Koppenburg requested to merge pkoppenb-DecayTreeFitterAlg into master

Test for DecayTreeFitterAlg and subsequent functors. Based on Phys!970 (merged) and Rec!2504 (merged)

Edited by Patrick Koppenburg

Merge request reports

Merge request pipeline #3275001 passed

Stage: test
    Stage: .post

      Merge request pipeline passed for 067a3745

      Merged by Patrick KoppenburgPatrick Koppenburg 3 years ago (Nov 24, 2021 7:32am UTC)

      Merge details

      • Changes merged into with 73bd2a95.
      • Deleted the source branch.

      Pipeline #3277856 passed

      Stage: test

        Pipeline passed for 73bd2a95 on master

        Activity

        Filter activity
        • Approvals
        • Assignees & reviewers
        • Comments (from bots)
        • Comments (from users)
        • Commits & branches
        • Edits
        • Labels
        • Lock status
        • Mentions
        • Merge request status
        • Tracking
      • Patrick Koppenburg marked this merge request as draft

        marked this merge request as draft

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        • ad4781b2 - Add DTF functor --- does not work yet

        Compare with previous version

      • Patrick Koppenburg mentioned in merge request Rec!2504 (merged)

        mentioned in merge request Rec!2504 (merged)

      • Patrick Koppenburg added 11 commits

        added 11 commits

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        • b237c419 - This cofnigures but DecayTreeFitterAlg crashes

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        • a315c3c2 - Updated example. Still not working.

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg mentioned in issue Analysis#17 (closed)

        mentioned in issue Analysis#17 (closed)

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg removed lhcb-run3-cleanup label

        removed lhcb-run3-cleanup label

      • Patrick Koppenburg added 11 commits

        added 11 commits

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg added 5 commits

        added 5 commits

        Compare with previous version

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        With Rec@8f240165 I get

        FunctorFactory                         INFO Cache miss for functor: ::Functors::Adapters::ParticleMapper<>( /* TES location of relation table */ std::string{"/Event/DecayTreeFitterAlg/OutputRelations"} ), now trying cling with headers [<string>, Event/Particle.h, Functors/Adapters.h]
input_line_459:9:176: error: no member named 'ParticleMapper' in namespace 'Functors::Adapters'
std::unique_ptr<Functors::AnyFunctor> functor_0xdacc4462ac30ad0e_cling() { return std::make_unique<Functors::Functor<std::any (LHCb::Particle const&)>>( ::Functors::Adapters::ParticleMapper<>( /* TES location of relation table */ std::string{"/Event/DecayTreeFitterAlg/OutputRelations"} ) ); }
                                                                                                                                                         ~~~~~~~~~~~~~~~~~~~~~~^
input_line_459:9:191: error: expected expression
std::unique_ptr<Functors::AnyFunctor> functor_0xdacc4462ac30ad0e_cling() { return std::make_unique<Functors::Functor<std::any (LHCb::Particle const&)>>( ::Functors::Adapters::ParticleMapper<>( /* TES location of relation table */ std::string{"/Event/DecayTreeFitterAlg/OutputRelations"} ) ); }
                                                                                                                                                                                              ^
FunctorFactory                         INFO Code we attempted to JIT compile was:
#pragma cling optimize(3)
#pragma cling load("TrackKernel")
#ifndef GAUDI_V20_COMPAT
#define GAUDI_V20_COMPAT
#endif
#include <string>
#include "Event/Particle.h"
#include "Functors/Adapters.h"
std::unique_ptr<Functors::AnyFunctor> functor_0xdacc4462ac30ad0e_cling() { return std::make_unique<Functors::Functor<std::any (LHCb::Particle const&)>>( ::Functors::Adapters::ParticleMapper<>( /* TES location of relation table */ std::string{"/Event/DecayTreeFitterAlg/OutputRelations"} ) ); }
auto const functor_0xdacc4462ac30ad0e_cling_factory_ptr = functor_0xdacc4462ac30ad0e_cling;
DimuonsTuple                          FATAL  Exception with tag=FunctorFactory::get<Functors::Functor<std::any (LHCb::Particle const&)>> is caught

        The good news is that is caught the TES location /Event/DecayTreeFitterAlg/OutputRelations. But then it doesn't like the rest.

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        Adding @graven to participants

      • Patrick Koppenburg requested review from @graven

        requested review from @graven

      • Patrick Koppenburg added 3 commits

        added 3 commits

        Compare with previous version

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        ... I think it's a git glitch. Trying again.

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        Indeed. It was a red herring. The DaVinci test initialises correctly with Rec!2504 (1fd0e12a)

        Edited by Patrick Koppenburg
      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        Success! I have a refitted PT in the tuple!

        DTF_pt

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        Now the next step is to generalise with any functor.

      • Eduardo Rodrigues
        Eduardo Rodrigues @erodrigu ·
        Maintainer

        Really nice :thumbsup:!

      • Patrick Koppenburg added 1 commit

        added 1 commit

        Compare with previous version

      • L
        Laurent Dufour @ldufour ·
        Developer

        Nice work! Does the test include the results for all of the different PVs in the refitting? It's not clear to me whether this variable is still an array :-) (and how do you access the DTF chi2?)

        Edited by Laurent Dufour
      • Eduardo Rodrigues resolved all threads

        resolved all threads

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        That will all come. Next I need to access PT via the functor. Once I can do that I can access any property of the refitted particle, and DTF chi2 can be one of them.

      • Eduardo Rodrigues
        Eduardo Rodrigues @erodrigu ·
        Maintainer

        @pkoppenb, so you're not accessing your DTF pT via a/the functor? Got me perplexed. But I can wait for the magic to be displayed soon ... looking really forward to it :-).

      • Patrick Koppenburg
        Patrick Koppenburg @pkoppenb ·
        Author Maintainer

        Not yet. We did a design session at the whiteboard with @graven. We decided to factorise the problem into two steps:

        1. Make a functor that accesses the related particle and returns pt
        2. Replace pt by a functor PT, or any functor you like.

        (Now I am surprised Gerhard didn't mention "divide and conquer" when suggesting that :-) )

        Step 1 is complete. Next week I'll get to step 2. It was useful to split the problem as it identified that we need two functions. Now I need to template them by (or ideally only the outer one).

        One problem is that TestFunctors.cpp in Rec has no access to the Gaudi TES and this cannot really test the functor, just make sure it complies. Hence we need the chain to work up to DaVini to be sure it's OK.

        Edited by Patrick Koppenburg
      • Eduardo Rodrigues
        Eduardo Rodrigues @erodrigu ·
        Maintainer

        Sounds good to me. I look forward to seeing a draft since point 2. sounds like a generic adaptor, as it were.

      • Loading
      • Loading
      • Loading
      • Loading
      • Loading
      • Loading
      • Loading
      • Loading
      • Loading
      • Loading
        • Please register or sign in to reply