Axel Donath

@StFroese Just looking at images, this is a very tricky case. There is clearly not enough statistics to provide any stable background estimate. Typically this is handled with some pre-process...

> I'm still a bit confused. Isn't this line here exactly allowing interpolation between energies, which should not be allowed by default? Yes, this is the intended behavior. It should...

Thanks @HealthyPear! The LHASSO catalog has also already been added to Gammapy: https://docs.gammapy.org/1.3/api/gammapy.catalog.SourceCatalog1LHAASO.html#gammapy.catalog.SourceCatalog1LHAASO I would argue that adding full catalog data is not really in scope for gamma-cat right now,...

Thanks a lot @registerrier this is important work! How do we find all the relevant places? Raising an error in `self.quantity` and running all the tests?

Thanks @registerrier! I just wanted to add some context to this change: the reason why we decided to split the datasets was indeed performance. So evaluating a set of smaller...

I agree "CC BY-NC-ND" is mostly appropriate. We could however think about allowing derivatives of the data? But that depends on the initial license of the data as well...

@bkhelifi Hm, but does this mean that people can never publish data results from a small study they made using Gammapy data?

The licence forbids "derivatives":  Does it mean if somebody computes flux points from the HESS DL3 DR1, they are not allowed to...

Refactoring and simplifying tests will be a big topic for after v1.0. Re-labelling now...

Thanks @QRemy, the statistical approach to handle the ULs looks reasonable to me. I was wonderign whether we can maybe find a more descriptive name. I think even "chi2-marginalize-ul" would...