Now for the very first time, I’m trying to see how RDataFrame::Vary could be used in an “ATLAS analysis”. I may very well not be the first one to try this, though I haven’t seen any presentations in the experiment yet on people having done this successfully yet.
I’m embarrassed to say how long it took me to figure out exactly how I should implement some functors that could be used with RDF::Vary...). But I think by now I managed.
The reason I’m opening this thread is because I found the documentation of the support for systematic variations in RDF a bit lacking. I gathered by now that this is still an experimental feature in the code, but the documentation is not even clear about that. For instance it took me a moment to find that the code example in ROOT: ROOT::RDF::RInterface< Proxied, DataSource > Class Template Reference should rather look like:
auto hx = ROOT::RDF::Experimental::VariationsFor(nominal_hx);
Since even in the master branch that function is still in the Experimental namespace.
But don’t understand my complaints too negatively. I really appreciate the work on this RDF feature. I just wonder if more discussion between ATLAS and the ROOT team could benefit this code.
ROOT Version: 6.28/10 Platform: Linux Compiler: GCC 13