Crash when extracting flux points with a CompoundSpectralModel

Gammapy version v0.19

Bug description If a compound model is used for the FluxPointsEstimator, an error occurs in FluxEstimator.get_scale_model() because a CompoundSpectralModel has no norm or amplitude attribute.

Expected behavior Using a CompoundSpectralModel should work.

The issue is caused by the mechanism introduced to scale the norm according to the input parameter. It does not work in general for a CompoundSpectralModel which might have different amplitude and norm parameters. So that determining the actual range a priori might be ambiguous.

Note that we miss a test for flux points extraction with a compound model.

To Reproduce

from gammapy.modeling.models import EBLAbsorptionNormSpectralModel, PowerLawSpectralModel, SkyModel
from gammapy.estimators.flux import FluxEstimator

pl = PowerLawSpectralModel()
ebl = EBLAbsorptionNormSpectralModel.read_builtin(reference='finke')
model=SkyModel(spectral_model=pl*ebl)

fe = FluxEstimator()
fe.get_scale_model([model])

Other information Any other information you think will be useful for us to fix the issue can go here.

[QRemy]

In principle the scaled parameter does not have to be one of the existing parameters as it is already defined by the ScaleModel so in the get_scale_model method you pointed we could have :

        scaled_parameter = None
        if "amplitude" in ref_model.parameters.names:
            scaled_parameter = ref_model.parameters["amplitude"]
        elif "norm" in ref_model.parameters.names:
            scaled_parameter = ref_model.parameters["norm"]
        if scaled_parameter is not None:
            scale_model.norm = self._set_norm_parameter(scale_model.norm, scaled_parameter)

[adonath]

Users now have to define the is_norm property, when a SpectralModel is used with a SkyModel. See https://github.com/gammapy/gammapy/blob/master/gammapy/modeling/models/cube.py#L75 The example for the custom model was adapted accordingly: https://docs.gammapy.org/dev/tutorials/api/models.html#Implementing-a-custom-model