Dan Foreman-Mackey

Which link is that? Here are the docs: https://george.readthedocs.io/en/latest/user/quickstart/#from-source

I don't have any plans for that. It's possible that you could find your way around the C++ source which lives here: https://github.com/dfm/george/tree/main/src/george/include/george but it wasn't designed for public consumption...

Good question! The priors returned by george are not properly normalized so you'd be better off doing it yourself.

This means that there is a compiler error when you try to compile. Take a look at what that error is and see if you can fix it.

Yeah - that interface is all a little janky! All of these suggestions are good. I don't have a huge amount of time for george maintenance these days, but I'd...

Awesome!! I think the easiest first step would be the improved error message. What do you think?

Thanks! This link if from the old version of the docs (the new ones can be found here: http://george.readthedocs.io/en/latest/). You can find the old examples here: https://github.com/dfm/george/blob/b2e4150cc1bd0c06c1cb323396731ff156ef4c7a/docs/_code/

I can reproduce the original error is related to the fact that with recent versions of numpy(maybe?) `-fno-exceptions` is included as a gcc compile argument. I can get this to...

The conceptual problem with this approach is that PyMC needs to work on a differentiable logp function, with support everywhere: (-inf, inf) in all the parameters. This prior doesn't have...

Working out that reparameterization would definitely be the best way to go about this, if possible. I haven't looked into this, but it would be cool to see! And yeah,...