FFNS nf=5 NLO grids

[giacomomagni]

Addressing #66

[Radonirinaunimi]

So the new fits have been re-organized here. In particular, the fit report for the new theory (6998304) is available in this link.

Since this fit already includes the $x$-extension down to $10^{-3}$ the only loose end is the double-counting in the matching. I've already ran a fit in which these double-counting are removed in the matching but it would be good to sill have both $F_{2,3}$ and $d\sigma$ observables by shifting the cross-sections.

https://github.com/NNPDF/nnusf/blob/7d201b0430496c873fbd9870b338997187bb4dfd/src/nnusf/data/matching_grids.py#L99-L103

@giacomomagni Could you please re-generate the double diff. cross section inputs (as the SF observables will be unchanged)?

PS: A shift of $10~\mathrm{GeV}^2$ should be consistent with the current spacing in https://github.com/NNPDF/nnusf/blob/ffns5_nlo_grids/commondata/matching-grids.yml (?)

[giacomomagni]

@giacomomagni Could you please re-generate the double diff. cross section inputs (as the SF observables will be unchanged)?

PS: A shift of 10 GeV2 should be consistent with the current spacing in https://github.com/NNPDF/nnusf/blob/ffns5_nlo_grids/commondata/matching-grids.yml (?)

okay, let me give a try.

[Radonirinaunimi]

Based on the results, this should now be merged. If further changes are needed this can be addressed in another PR.