Closed Radonirinaunimi closed 2 years ago
@giacomomagni I think that we can try with this as a start. The real datapoints will then be cut off at $QI^2 = 28~\rm{GeV}^2$ from which the matching grids will start. For $Q{II}^2$, Juan suggested to take at least $Q{II}^2 = 900~\rm{GeV}^2$, however, I think we should check first if the smaller values of $Q{II}^2$ we could fit both w/ and w/o input scaling. Also, given that some of the real datapoints will be removed (for NUTEV about 500 will be cut off) I refined a bit the grids.
@giacomomagni, could you please generate the new grids with the latest commit. So basically the points are now different for each experiment and they correspond exactly to the $x$-points in the dataset. While doing so, could you also double check if something has been broken?
Okay @Radonirinaunimi. I'll update the grids.
The kinematics per experiments are now fixed, some notes for future reference:
This should be merged now and once https://github.com/NNPDF/nnusf/pull/41 is also merged (which I'll do in a while) then we can run the long-awaited fits.
Changes the kinematic grids used to generate the Yadism matching data.