GammaLib

The H.E.S.S. community uses effective area components that have a theta2 cut applied (see https://gamma-astro-data-formats.readthedocs.io/en/latest/irfs/point_like/index.html). It should be possible to use these IRFs in csphagen (and ultimately in GCTAOnOffObservation without recomputing the PSF containment fraction.

Such an IRF can be recognised using the following characteristics:

• HDUCLAS3 keyword is POINT-LIKE
• Presence of RAD_MAX keyword

So if a POINT-LIKE effective area is encountered, the code should not correct the ARF with the PSF containment fraction. The code should verify that the RAD_MAX keyword is consistent with the radius of the On region, and if this is not the case, throw an exception. Ideally, the code should also verify consistency with the On region centre, but it seems that this information is not available. Also, this logic only works for circular On regions, and an exception should be thrown if the On region is not circular. And finally, the code only works for point sources, so an exception should be thrown in case that a non point-source model is detected.

All the checking could be implemented in the GCTAOnOffObservation::set method, and if everything is okay, a flag could be passed to GCTAOnOffObservation::compute_arf to switch off the theta2 cut. This should probably be reported in the ARF for diagnostics. A possibility is to add a RAD_MAX keyword in the ARF file in case that the IRF had such a parameter. This means that the GArf class needs to be modified.

Reading of the HDUCLAS3 and RAD_MAX keywords needs to be implemented at the GCTAAeff2D level. From there they can be accessed by the GCTAOnOffObservation::set method.