Jet Energy Scale systematics can be split by sources and grouped

[GaelTouquet]

For the H2TauTau analysis, JES is split by sources and grouped as is explained here.
I have implemented the possibility of doing so in the jet recalibrator. Maybe this could be useful in other analysis too.
This has not been fully validated yet but will be within a week, I will add a comment here when it will be so.

[GaelTouquet]

I have committed a fix for the requested change, but I haven't completely synced with other groups on the grouped systematics yet. It might not come from this code but from the analysis part, it is still under investigation. As said before, I will sync this ASAP and confirm here that it gives a satisfactory result.

[GaelTouquet]

Just a quick update : I have checked the code with an other group and everything seems to be alright but there is still disagreement on the synchronization of a test sample. Someone from the group agreed to do some scrutinized parallel testing, but he will only be available to do so beginning of next week.
Sorry to keep you waiting

[GaelTouquet]

We have reached synchronization! The group I was syncing with had overwritten pointers...
Although we haven't decided if we should keep the len(sources) == 1 case as was before, leading to the inversion of up and down shifts in this case.
I suggest to merge this the way it is and I will find a workaround in our analysis code if needed.
Thank you for your patience.