Making `zea.data.generate_zea_dataset` more robust

[tristan-deep]

Currently we have an example on how to generate a zea data file here.

Some things that are missing and necessary to make this as robust as possible for various amounts of data acquisition setups:

• Create a test that also processes the data file, mainly beamforming, to always ensure the dataformat is correct.
• Check generate_zea_dataset for default arguments and optionals and if in these cases the above tests still run.
• For instance polar_angles is not mandatory, but without it beamforming will fail. Should be fixed.

[wesselvannierop]

I think this issue touches on something deeper, that we could choose to tackle in the corresponding PR. This is that the code for the data format (i.e. scan parameters) is a bit scattered. We have zea/data/data_format.py, zea/data/file.py, zea/internal/checks.py, zea/scan.py where all of them are handling ultrasound (scan) parameters, their relation, checking them, validating etc. Maybe we can clean that up a bit, such that you can check the scan parameters from an exisiting file, during the creation of the file, and as a scan object, all using the same logic.

[wesselvannierop]

Note that this will be File.create with the openh-rf updates.