Draft: Make pseudo data handling in JetUncertainties tool nicer
Make pseudo data handling in JetUncertainties
tool nicer. It still requires two sets of tools, but one now only sets up JER systematics and suffixes them with PseudoData
.
CP algorithms have been updated using AnalysisTop notation, e.g. setting JEROption = "Full_PseudoData"
.
Open questions:
- I now rely on string manipulations only on
CP::SystematicVariation
as I assume original names are needed for histogram loading. - How to skip JES systematics reliably?
- If both tools are configured, both tools are called as nominal running seems to be a void operation, is this problematic?
- A related question for @krumnack: how to check in a proper way that the
CP::SystematicSet
that we get from the loop is affecting the tool? At the moment onlyCP::SystematicVariation
s can be checked in the interface usingisAffectedBySystematic
.
Tagging many people: @jburr for CP algs, @mswiatlo, @camacho, @schaarsc and @sawyer for jets, @omajersk and @mrimoldi as they may be interested in that for AnalysisTop/SUSYTools (feel free to tag more people if I have forgotten some).
Keeping Draft for now to allow some discussion.