Skip to content
Snippets Groups Projects

Substitutions in DTFDict

Closed Alessio Piucci requested to merge apiucci-SubstituteInDTFDict into master
1 unresolved thread

Implementation of Substitutions in DTFDict, taken from the TupleToolDecayTreeFitter.

Edited by Alessio Piucci

Merge request reports

Loading
Loading

Activity

Filter activity
  • Approvals
  • Assignees & reviewers
  • Comments (from bots)
  • Comments (from users)
  • Commits & branches
  • Edits
  • Labels
  • Lock status
  • Mentions
  • Merge request status
  • Tracking
55 60 declareProperty
56 61 ( "constrainToOriginVertex",m_usePV,
57 62 "flag to switch on PV contraint");
63
64 declareProperty( "Substitutions", m_map,
65 "PID-substitutions : { ' decay-component' : 'new-pid' }" ) ;
  • Have you discussed this addition with @sneubert? Adding him for information and feedback. Do you have any test for this, which would make sense to add to the nightlies suite? My other comment is about the branch. It would make sense to have this in the 2018-patches branch and then propagated to master. You committed to master, though.

  • Hi Eduardo,

    I am aware of the feature Alessio has implemented. We discussed it and it's a very nice addition. Cheers, Sebastian

  • Thanks for the quick confirmation.

  • Edited by Software for LHCb
  • Alessio Piucci added 1 commit

    added 1 commit

    Compare with previous version

  • @erodrigu : Unfortunately I don't have any test to be added to the nightlies right now, I've only produced some ntuples to test the feature. But I can have a look how to implement a test, if required.

    About the branch: sure, I could apply the MR to the 2018-patches. However there's no opportunity to change branch on the fly anymore, as far I can see. Should I close and reopen the MR on that branch?

  • Hi @apiucci, having a test would be great. It's so much better to have tested code ;-). We are far from perfect so it would make sense to work better when adding code. You could also work with @sneubert to make sure the test tests the full algo, actually.Typically we get code to the 2018-patches and I then cherry-pick to master. This is very clean. If you can easily commit this same thing to 2018-patches then thanks. You would then close this one.

  • Hi @erodrigu , sure there's no problem. I've written a simple test that I will commit to the next MR pointing to the 2018-patches branch. Do we have already a test sample that we can use to test the algos?

  • Brilliant, thanks a lot! Hmm, test files. There are a bunch of files in the TestDB. In principle any data should be fine. It would be a matter of choosing a recent file, say from run II?

  • I found a test file in the TestDB which is suited for the test, but is 2012. I hope that it's still fine. If so, I will close this MR and open the new one on the correct branch.

  • I reckon that's fine. After all, you do not want to check physics here but rather the correct manipulation of some algorithm/tool. Go ahead. Thanks again!

  • Alessio Piucci mentioned in merge request !253 (merged)

    mentioned in merge request !253 (merged)

  • Please register or sign in to reply
    Loading