cms-jet/jme-validation

jme-validation's Issues

Access to tree.Jet in src/definitions.py: possible source of errors

Calling tree.Jet in src/definitions.py (this line) will access different jet collections depending on the input sample. The Jet collection is CHS in Run2 samples and PUPPI in Run3 samples. This may cause users to retrieve jet collections inconsistently.

Things can get more complicated when accessing non-standard flavors of NanoAOD, e.g. JMENano where both PUPPI and CHS are available, or custom JMENano with additional ABC collections. A way to deal with this (more elegant than hard-coding the collections to be retrieved) may be needed.

'plot_level' not used in DYModule.py

While QCDModule.py uses the 'plot_level' option to decide whether or not to plot responses (see here), the same does not happen in DYModule.py (see here), i.e., DYModule.py plots responses regardless of the printing level.

On a side note, some plot naming may be wrong: for example here we apply noSel selection but call the plot hasTwoSFLeptons, which sounds inconsistent.

Name change for rho variable between Run2 and Run3 NanoAODs

The README currently recommends to do

sed -i 's/"fixedGridRhoFastjetAll"/"Rho_fixedGridRhoFastjetAll"/g' bamboo/bamboo/analysisutils.py

in order to run on Run3 NanoAOD, where the ρ variable is called Rho_fixedGridRhoFastjetAll. This clearly makes the code crash when attempting to read Run2 samples, which still have the old naming.
This needs either to be treated in the code or harmonized by future NanoAOD versions (maybe with coherent naming?)