Bruno Khélifi

Dear all, With this PR, the processing of HAWC data will be really possible. Thus, if all agree, I propose to you to make some update of the doc in...

@registerrier (2nd comment): I agree, even if it seems that in this exemple we are not in this case @AtreyeeS and @adonath : when having a very large RoI, its...

Agree. But we only encourage, not force... This is why I found the bug

With the current code, the example returns: `Effective area is all zero at [85d37m58.8s 22d00m50.4s]. No safe energy band can be defined for the dataset 'obs-23592': setting `mask_safe` to all...

Maybe, one should involve the Provenance people in such discussion... Indeed, a duplication of metadata or of methods should be avoid. I trigger then Mathieu Servillat : @mservillat .

This is associated to the PR #3970 .

So, removing the contributing section from the readme.md, and add just a link towards this new file into the readme?

If I understand correctly, the proposal is to remove https://github.com/gammapy/gammapy/blob/master/gammapy/modeling/models/cube.py#L86 AND to check in the FluxPointEstimator if the spectral model has several free norm??

In the `FluxEstimator`, the first parameter is always the one of the `ScaleSpectralModel` that is created internally by the estimator, the parameters of the other models (in particular the ones...

So, the idea is to create a new PR for `FluxEstimator.get_scale_model` containing a check on the number of non-frozen 'norm' parameters of the reference model is greater to 1... Is...