Remove overlaping leptons in data
The main focus here is to extend the (normally truth-based) overlap removal so that we can remove reconstructed leptons from data.
I had to make a few changes to make this work:
- I gave the
JetMatcherAlg
the ability to write the matched particle as a link on the jet. This is used in the overlap removal now. - Since I match to both muons and electrons, I changed the
JetMatcher
block specification a bit so that it can take lists of inputs.- I also removed the requirement that some variable be copied to the target jet, since the matched particle is good enough.
-
❗ I fixed a bug in the truth configuration in the dumper: we were decorating every jet with truth information even if people didn't ask for it.
This was built with help from @jackersc, but I think it's about ready to be merged now. Jona, I merged the config file we were validating this with with the existing data file. The difference was pretty small, and I think we're the main users of the data pipeline. I thought better to avoid creating too many orphan config files.
Note that the lepton overlap cut seems pretty harsh at the moment: in a sample of 10 events I went from 17 jets to 6 jets, and that with a 15 GeV requirement on the reconstructed leptons. We could study this in a whole lot more detail by running on simulation and seeing how often the removed jets are actually leptons, and adding more stringent lepton ID would probably help a lot. But I think this is good for a first go.
Also tagging @npond, hopefully he can have a look.