Update 06.28.2018 -- Run 9 embedding: using phi vs. delta-phi and particle-lvl. triggers in matching

Yesterday, I investigated a bug in our matching code and the impact of a max 'qT' cut:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-06272018-run-9-embedding-applying-max-qt-cut

Today I'm investigating the impact of using phi vs. delta-phi in matching and requiring a trigger at the particle level.  Up until now, I had only been paying attention to the reconstruction efficiency up to 'pTjet ~ 9 GeV/c', which had looked reasonable.  See, for instance:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-06052018-run-9-embedding-new-matching-algorithm

In that post, I was only requiring a trigger at the detector level but not at the particle level.  So there were events with a "trigger" at the detector level but no trigger at the particle level.  Regardless, in all cases I required that we only consider "recoil jets" at both the particle and detector level.  This resulted in a sudden drop in reconstruction efficiency at 'pTjet = 8 GeV/c' or so.

Requiring a trigger be present at the particle level (i.e. there is a charged hadron with 'pT = (9, 20) GeV/c' and '|eta| < 0.9') improves the reconstruction efficiency across the board and especially around 'pTjet = 9 GeV/c'.  Moreover, I was curious about whether or not matching in raw phi space or delta-phi space made a difference.  Judging from the reconstruction efficiency, it doesn't seem to make a big difference; though the efficiency for matching in delta-phi space is systematically above that for matching in phi space above 'pTjet = 9 GeV/c'.  In light of this, I would say requiring a particle level trigger and matching in delta-phi space is the ideal method in this case.

Below is the delta-r distributions ('dR = sqrt((phiMatch - phiTruth)^2 + (etaMatch - etaTruth)^2)' or 'dR = sqrt((deltaPhiMatch - deltaPhiTruth)^2 + (etaMatch - etaTruth)^2)') for the same configurations:

And for completeness, here's how the delta-r distributions vary with partonic 'pT' (requiring a particle level trigger and matching in delta-phi space):

I've attached the relevant ROOT files as a zip file.

Note: in all cases, when matching, phi and delta-phi were restricted to the interval (0, 2pi).