Add support for variable scoping about formulate HOT 6 OPEN

Yes, I definitely want to add support for this!

I don't use files like this very often, would you be able to share an example file? You can send it to me privately if you'd prefer by emailing me at christopher.burr with the domain being cern.ch.

from formulate.

The samples I'm developing against are a bit private atm I'm afraid, but I think the mc10event.root file in uproot's testing samples is a good equivalent for what I'm looking at:
https://github.com/scikit-hep/uproot/blob/master/tests/samples/mc10events.root. Would that work for you?

from formulate.

Thanks for file (and the delimitedList suggestion), I've added basic support now in v0.0.8. The example you gave now looks like this:

$ python -m formulate --from-numexpr 'branch.sub_branch < 4' --variables
branch.sub_branch

I don't work with files like this very often so please open issues if you run into more limitations. 😄

from formulate.

Same comment made at #22 - example above worth having in the README. Maybe use the file above for a test?

from formulate.

Thanks for such a speedy implementation, Chris!

from formulate.

Hi I have a similar problem with with variables that have dot "." inside. See

As you can see, this is done with the last version of formulate, i.e v0.08.

The variables that bring problem are the isolation variables header_0.30_cc_asy_PT where 0.30 stands for the cone radius, this variables is automatically created by a LHCb tupletool.

An additional concern that I have, is how then the numpexpr conversion will be done?

from formulate.