Sergey Koposov

I think the right way is to put the deprecation warning and make the n_effective a no-op in sample_initial() That also means we need to deprecate n_effective_init in run_nested() of...

Hi @edbennett, Sorry, I've been traveling, and didn't have much time. I'll try to review this in next couple of days.

Thanks for the patch, I've looked at it now. The dynamic nested sampling n_effective is indeed used and should not be deprecated. I've left a small comment on that for...

Okay, to avoid delaying further, I'll merge it now. Thanks for the change! To be honest, we probably want to restructure this bit, i.e. completely get rid of this https://github.com/edbennett/dynesty/blob/4e0594eac0f2fdd224ab8b79dd348813a842e467/py/dynesty/sampler.py#L683...

The simplest way is to use a generator https://dynesty.readthedocs.io/en/latest/quickstart.html?highlight=generator#running-externally while pickling periodically the sampler. The HDF5 history won't help here much though (as hdf5 history only saves the function evaluations,...

For people observing this issue, I have a test implementation of the resuming interrupted runs in PR #386 . Any feedback on the implementation/interface would be helpful.

The #386 has been merged, so this issue is resolved.

Thanks for the PR. In my opinion I would prefer to start merging just the ensenble sampling part and leave outside the axis based proposals and chi-proposals (unless there is...

I have spent some time looking and thinking about this proposal, and my current thinking (on top of what I said previously) is that * we should keep the existing...

I have started a test branch for possibly way of implementing this change https://github.com/joshspeagle/dynesty/tree/extra_info that will still need a lot of work and it may still not be the right...