add g4_reg nc

[RoyStegeman]

Includes polarized structure function $g_4$ at NLO

Reference for NNLO (eq. 19): https://arxiv.org/pdf/2210.12014.pdf

[RoyStegeman]

@felixhekhorn if I run the benchmark I get only 0.0 as ouput for the apfel g4total. Is this something you understand?

Please ignore unrelated aspects of this PR, there's a reason I put it as draft ;)

[Radonirinaunimi]

LGTM!

Just to mention that now all the references to the de Florian paper should be updated:

yadism/src/yadism/coefficient_functions/light/g1_nc.py

Line 22 in 44341a3

:cite:`deFlorian-light-nlo-pol`.

yadism/src/yadism/coefficient_functions/light/g1_nc.py

Line 37 in 44341a3

For a simpler expression, refer also to :eqref:`9` of :cite:`deFlorian-light-nlo-pol`.

[RoyStegeman]

That's the nlo paper, this is the nnlo paper.

Or do you mean replace the nlo with the nnlo paper? I don't think that matters very much if both are correct.

[Radonirinaunimi]

That's the nlo paper, this is the nnlo paper.

Or do you mean replace the nlo with the nnlo paper, I don't think that matters very much if both are correct.

Yes, I meant replace. But as you say it doesn't matter that much, at least unless we use the equivalence relation everywhere.

So, please feel free to merge this into #173.