Flat TTree's are pretty common. This would enable the system to run against them.
Deal with flat ntuples that contain just numbers, and it could be you get some number of things per event (e.g. jet pt's).
Do not try to make a "jet" object - leave that for a new problem when this one works.
A lot of regular C++ code rendering could be reused for this. This will be a totally new backend and executor, however. Put root in docker just to be safe that we use a known and well understood version of ROOT. If possible.
How should the back end get coded? RDataFrame would be awesome.
Flat TTree's are pretty common. This would enable the system to run against them.
A lot of regular C++ code rendering could be reused for this. This will be a totally new backend and executor, however. Put root in docker just to be safe that we use a known and well understood version of ROOT. If possible.
How should the back end get coded? RDataFrame would be awesome.