Improve particle ID handling

[cschwan]

We should improve the particle ID handling in PineAPPL. Some things that should be done:

• Make particle ID bases transparent. See also Support evolution basis FkTables #75 (comment).
• Starting with commit 4e40743 we're implicitly assuming that the PDFs accept only PDG MC particle IDs. Shall we support also PDFs in the evolution basis?

[felixhekhorn]

* [ ]  Starting with commit [4e40743](https://github.com/N3PDF/pineappl/commit/4e40743929483f9e670069238e7368b77634ba76) we're implicitly assuming that the PDFs accept only PDG MC particle IDs. Shall we support also PDFs in the evolution basis?

I think we can rely on the input beeing lhapdf-like, meaning only accepting PDG MC PIDs - is that what you mean?

[cschwan]

* [ ]  Starting with commit [4e40743](https://github.com/N3PDF/pineappl/commit/4e40743929483f9e670069238e7368b77634ba76) we're implicitly assuming that the PDFs accept only PDG MC particle IDs. Shall we support also PDFs in the evolution basis?

I think we can rely on the input beeing lhapdf-like, meaning only accepting PDG MC PIDs - is that what you mean?

That's what we're doing right now, but I wonder whether this is a restriction. @scarlehoff When we ask the NNPDF library for a PDF, do we get it in the flavour basis, evolution basis, or can we ask for it?

[scarlehoff]

You can ask for the basis you want when using vp.

It's only the convolution that has this restriction of the "evolution basis" (which, more than a restriction, as we have talked, is a convenience of not mixing basis)

[felixhekhorn]

But I guess for vp you don't get a callable xfxQ2(pid,x,Q2), but just a bunch of numbers, right? (because the 200=Singlet is an invention of us ...)

[scarlehoff]

But we can create that callable easily (having access to such a callable is one of the things I've done. At the moment only accepts the lhapdf MC PIDs because I wanted it to be lhapdf-palatable but adding the rotation if needed is trivial.