current effective exponents implementation isn't flexible enough

[wilsonmr]

CC: @scarlehoff

I helped implement the plotting of effective exponents and next effective exponents in validphys and I was just looking but at the moment it won't be very helpful because you can't specify a different basis than the fit basis.. So this needs to be added, it shouldn't be too much effort

Also as a side note the flavour basis is available in validphys.. line 309 of pdfbases.py and we use it in the fit comparison reports, so I don't think there is an issue with stability of this

[wilsonmr]

the changes are essentially at the level of the *_internal functions in eff_exponents.py where we want to be able to set the basis ourselves. We can discuss at code call then I can probably take care of this later

[Zaharid]

Just to say that indeed the relevant pieces of vp are pdfbases, which should be the only place in the whole code that deals with rotations (in particular see #559) and eff_exponents, which deals with well, effective exponents.

[juanrojochacon]

I have self-assigned to this PR in case I can check with the implementation and checking of the flavour basis!

[juanrojochacon]

Just to say that indeed the relevant pieces of vp are pdfbases, which should be the only place in the whole code that deals with rotations (in particular see #559) and eff_exponents, which deals with well, effective exponents.

Very nice. In particular I think we all agree that pdf basis and eff_exponents should be decoupled, and that one can evaluate effective exponents in any PDF bases, and in particular in other bases as compared to those used in the fitting

[scarlehoff]

Just adding an item to the wishlist, it would be useful to have some documentation on how the different flavour basis should be used in the runcard. One can make an educated guess by looking at the code but that's less than ideal.
@scarrazza just noticed that evolvefit is using the basis information from the runcard as well to do the evolution, what for? The LHAPDF grids are always in the flavour basis regarding of the runcard basis, right?

[juanrojochacon]

Hi @scarlehoff @scarlehoff yes the LHAPDF grids are always in the flavour basis. However the evolution is carried out in the evolution basis and then rotated back in the flavour basis. So one needs to have some memory of this rotation to be able to produce the LHAPDF grid files.