Scale variations

[felixhekhorn]

No description provided.

[alecandido]

• separate nonsinglet, singlet, gluon kernels&operators in order to apply scale variations (they apply differently to each of them)
• deny factorization scale variations (but not renormalization ones) for intrinsic (and so maybe matching)
• think about recombining s, ns, g (e.g. gVV and gAA, or FONLL bits) before or after blow-up (i.e. blow-up before or after splitting functions)
• give up on ren_to_fact and come back to xiR, since for intrinsic we need to deny xiF and xiF only
• for FONLL: we need to use fact sv in the 4 flavors since PDF are expressed in the 4 flavors (even if there are 3 bits, they are make up to behave like a single one)
• for Intrinsic FFNS: no fact sv, since the PDF is scale independent
• for Intrinsic FONLL: fact sv in 4 flavors scheme, since everything it is re-expressed in the 4 flavors scheme