nnpdf
nnpdf copied to clipboard
NNPDF
FPF data
I've regenerated some of the fktables and uploaded them to th 400 (i.e., if you download theory 400 right now you should be able to fit hte FPF data with this branch). I've also rebased to current master.
Some datasets are well fitted for some replicas while other aren't. I'll try doing a full fit to have a better look.
I've opened this as a dont-merge-PR to speed up the process of debugging possible problems.
As discussed, the default should be to have to total systematics $(E_\ell, E_H, \theta)$ + satistical uncertainties in the fit, and then a variation in which only statistical uncertainties are accounted. I therefore updated the $\rm{FASER}\nu 2$ files to account for the full systematic errors.
I am planning to compute everything for the other experiments (hopefully including the ekos and fks) by the end of today and then will ask you @scarlehoff to run your add_db.py script.
As for the commondata with statistical error only, I think @peterkrack is taking care of that.
(hopefully including the ekos and fks) by the end of today and then will ask you @scarlehoff to run your add_db.py script.
Ok! There's has been some problems in Milan so I don't have access to the computer there, but if you compute all fktables I should be able to run the fits normally.
What is the plan for this PR? I assume it was used for your recent FPF paper, but can it be closed or merge now?
What is the plan for this PR? I assume it was used for your recent FPF paper, but can it be closed or merge now?
Perhaps @scarlehoff has a preference (as initially he did not intend to merge this)?
Once the new commondata is done I'll port this data to the new format and then it can be merged. Don't want to merge more old commondatafiles when those are going to disappear "soon".