This is getting large enough that I almost feel like it should be organized as a class.

I am pretty sure this is not a comprehensive list of all the different kinds of flux units used in Astronomy. If we are going to make a class just based on this, it won't be complete. And to complicate matters, flux definition with the same name has different units in Astronomy and in Optical Engineering. Are we opening a can of worms here?

The way the equivalencies system is desined, I don't really see how a class would be helpful here - in the end it still has to look like a function.

What does make sense from the user perspective is to have this be a default equivalency for a Flux class. This was actually the primary use case that was in mind when the equivalencies system was designed. #1577 (or perhaps a follow-on PR built off that) should definitely implement that.

Perhaps I can explain: That's starting to be a lot of conversion functions, and every time this equivalency is used those functions have to be redefined with the outer function is called. I'm not sure how aggressively Python caches those. Right now it's not a big deal and I'm sort of off in premature optimization land, but it might make sense to define all those converters outside the actual spectral_density function. And I was thinking, in order to keep things organized, might as well make it a class. Something like:

class SpectralDensityEquivalencies(object):
    @staticmethod
    def converter_photlam_fla(x):
        return ...
    etc...

    def __call__(self):
        return [ equivalencies list...]

spectral_density = SpectralDensityEquivalencies()  # Basically a singleton

I should add: I wouldn't actually bother changing this right now. It's fine as it is. But something to keep in mind for the future maybe.

👍 Ah, I misunderstood. This is a good idea.

I'll do it if @eteq agrees.

Now I get it, yeah, that seems sensible, but I agree that should be a separate follow-on PR.