Inconsistency in relative coordinate definitions
We should standardise whether we compute jet.phi - thing.phi
versus thing.phi - jet.phi
. The same is true of deta
.
Currently this is not consistent
With respect to the jet:
With respect to the other object: