Axel Donath

> The stat_scan values in tests/spectrum/flux_points/binlike.fits are not compatible with the results. Remove? While this is true, it still might be the format that is used by fermipy. Not sure...

Test improvements will be part of the roadmap for v2.0

Thanks @maxnoe! While I don't have a strong opinion here I'd be happy to adopt the recommendation and put the SI brochure as a reference. This mostly concerns #3740...

Now one can actually do: ```python3 from gammapy.maps import axes axes.DEFAULT_LABEL_TEMPLATE = "{quantity} / {unit}" ``` Do you reuest anything else here @maxnoe?

Thanks @bkhelifi for the issue! I remember I have seen this in the past and the conclusion was that this is expected behavior: when plotting an integral flux quantity it...

@AtreyeeS We should probably even change `SpectralModel.plot()` to allow for passing energy axis. Such that the following works: ```python axis = flux_points.geom.axes["energy"] spectral_model.plot(energy_axis=axis, sed_type="flux") ```

@AtreyeeS Thanks for pointing to the discussion. I guess we wanted to keep the convenience of not having to define an axis just for plotting. However I think it would...

I just saw the corresponding FAQ entry: https://docs.kidger.site/equinox/faq/#typeerror-not-a-valid-jax-type It is explains how to adjust the filtering, however I would see the function as part of the "meta data" of the...

Thanks @patrick-kidger, for the clarification! I would guess the case of the callable not being a node is more common. But either behavior seems fine. Maybe adding one sentence to...

See https://github.com/patrick-kidger/equinox/pull/1009