Study reuse of puppi weights for Zinv
For the Zinv background estimation with AK8 jets (i.e. for the H(bb)+MET analysis), we are now using puppi. We remove photon or dilepton candidates from the PF candidate collection, and then the puppi weights are recalculated. We should try to use the default puppi weights included in miniAOD and see if the kinematic distributions come out the same. This would reduce the CPU usage when producing ntuples.
To test this, add the following line after the jet toolbox call in doZinvBkg.py:
process.puppi.useExistingWeights = cms.bool(True)