Someone asked me, "How did Kepler figure out when Mars was at opposition without having access to accurate clocks?

I wondered if he might have taken the midpoint of its retrograde loop. How close is that to the time of opposition? Neither of us knew.

So I wrote a program using PyEphem to calculate that. But along the way, I hit a snag in PyEphem (which turned out to be merely something missing in the documentation, easily fixed once I figured out what it was doing) and wondered if PyEphem was obsolete and maybe I should be using astropy. So I wrote a version in astropy.

Answer: No, stick with PyEphem. It's faster by a huge margin, and it gets more or less the right answer where astropy doesn't (I'm still not clear why not). Still, astropy may improve, and it's useful to have an example of how to use it, so I've kept it.

It turns out the author of PyEphem has a new library out called SkyField which is probably worth a look. I haven't tried it yet.

Oh, and if you're curious about the Kepler question? The PyEphem version currently says (if I got my calculations right) that the retrograde midpoint is roughly halfway between opposition and closest approach, and the difference between opposition and the midpoint is about 32 arcmin (half a degree). Would that have been close enough to help Kepler? I don't know. I'm passing the info on to my friend who's doing research for a talk and will let him figure that out.