Closed jonas-eschle closed 2 years ago
Base: 93.51% // Head: 93.05% // Decreases project coverage by -0.45%
:warning:
Coverage data is based on head (
d50fb77
) compared to base (528c2d4
). Patch coverage: 91.83% of modified lines in pull request are covered.
:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
This moves for asymptotic asimov dataset creation on the zfit binned PDFs and datasets.
Abstracting this away makes it easy to go for multidimensional asimov datasets (as zfit supports it natively) and fixes #55.
Also contains some other fixes.
I've though about the way to do it, whether we wanna replace the unbinned NLL (the
self.loss
) of the Asymptotic calculator with the Binned one, but I thought it's maybe better to keep it an implementation detail.What do you think @marinang (let me know if you're not available)