Dan Foreman-Mackey

This suggests that there's some sort of general problem with the Python installation, but I'm not sure what to suggest! It could have something to do with running within spyder...

Instead of using the `threads` argument, you should try the `pool` interface: [docs](https://emcee.readthedocs.io/en/stable/tutorials/parallel/#multiprocessing). The way you have it written, the multiprocessing pool is never being closed so you end up...

You can reparameterize to sample in an unconstrained 4-D space, e.g.: ``` theta_1 ~ N(0, 1), theta_2 ~ N(0, 1), theta_3 ~ N(0, 1), theta_4 ~ N(0, 1) ``` then...

Thanks for reporting. Are you setting your seed to be the same? In other words, are you sure that this is really being caused by the `pool` argument? If you...

Yes - this is a nice proposal and I'd be happy to review such a PR, but I'm not immediately sure how painful it would be to implement. If you're...

@cbeauc: Sorry about the slow response. I totally agree that it would make sense for `accepted` to be available on `State`. I think the reason for the current design is...

This isn't supported out of the box, but you should be able to arrange something like this by writing to the back end directly. You could try something like: ```python...

`emcee` will use a number of cores equal to half the number of walkers. So in your case, you might want to use 160 walkers to take advantage of the...

Yeah, I think it's right that you can't directly use multiprocessing for the model parallelization. I normally parallelize at a lower level (using numpy or hand-written C code), but that...

Thanks for reporting! You're absolutely right that there are several typos in the docs. The first example should read: ```python def log_prob(params): lp = log_prior(params) if not np.isfinite(lp): return -np.inf,...