Process partially a TTree with RDataFrame

[wiso]

Hello, is there a way to tell RDataFrame to process only the first 1000 entries in a TTree? Or to process the entries in the slice 1000-2000 (process 1000 entries skipping the first 1000)?

[eguiraud]

Hi,
you’re looking for Range. Note that Range is incompatible with multi-threading.

Cheers,
Enrico

[wiso]

Thanks! That’s what I needed. Sorry, I couldn’t find any example about it.

It’s a pity that it is incompatible with MT.

[eguiraud]

There is a tutorial here.

@wiso

It’s a pity that it is incompatible with MT.

I completely agree. Unfortunately I don’t know of a way to have multi-thread Range without introducing locks or synchronization points within RDataFrame’s event loop. Something that’s a bit more awkward but should work with multi-threading is adding a TEntryList to a TTree (or a list of TEntryLists in case of a TChain), and then passing the tree to RDataFrame’s constructor.