The source project of this merge request has been removed.
Instr met stat unc
- add correct stat unc on the Instr.MET : it now contains the reweighting uncertainty (results: http://homepage.iihe.ac.be/~hdelanno/SHEARS_PLOTS/plots_comparison_redIsWithoutReweightingUnc_BleuIsCorrectStatUnc/absoluteNumberOfEvents/ )
- correct a bug where we were using the inclusive reweghting function for the ee sample in all case. This is fixed now but it had little effects.
- add a check in the launchAnalysis script to verify if all files have been correctly processed before trying to merge them or run the dataMC script.
- change the final plots to have a content per bin instead of per GeV.
- update the reweighting to be only done with events <125GeV in MET, making the signal and the reweighting region independent (conceptually:= big difference; in practice=no difference)
- updated InstrMET weight files with those changes
- updating the lumi to match the full lumi of 2016
- added back the lines with no events in the Baobab list to be sure to "not forget we ran on those but notice that they contain no events in the golden json".
- and minor updates to make things better.
Edited by Hugo Delannoy