Jets can be recalibrated in nanoAOD-tools thanks to the JetRecalibrator.py.
Current modules use the recalibrator to propagate the effects to the MET:
in jetRecalib.py effect of recalibrated jets w/ pt > 15 GeV is propagated to MET: see here.
in jetmetUncertainties.py only effect of jet resolution is propagated to MET in MC: see here and here, while the effect of recalibrated jets seems to be neglected
I tried to reimplement the same logic that was implemented in heppy to calculate the T1-MET corrections and apply them to the raw MET, the current work-in-progress implementation is here (3.).
Below are a few checks I have performed to see the level of closure of these methods:
corrected MET with method 1. vs T1-corrected MET in heppy: pt, phi
corrected MET with method 3. vs T1-corrected MET in heppy: pt, phi
While method 3. has a better closure than method 1. the level of closure does not seem so satisfactory.
Note that T1-corrected MET in heppy closes well enough with out-of-the-box MET from nanoAOD: pt, phi
arizzi
commented
5 years ago
Jets can be recalibrated in nanoAOD-tools thanks to the JetRecalibrator.py.
Current modules use the recalibrator to propagate the effects to the MET:
I tried to reimplement the same logic that was implemented in heppy to calculate the T1-MET corrections and apply them to the raw MET, the current work-in-progress implementation is here (3.).
Below are a few checks I have performed to see the level of closure of these methods:
While method 3. has a better closure than method 1. the level of closure does not seem so satisfactory.
Note that T1-corrected MET in heppy closes well enough with out-of-the-box MET from nanoAOD: pt, phi