update error message if trying to get bounding ellipse of a single point

[segasai]

Helps #302

[coveralls]

Pull Request Test Coverage Report for Build 1148893105

• 153 of 167 (91.62%) changed or added relevant lines in 7 files are covered.
• 4 unchanged lines in 3 files lost coverage.
• Overall coverage increased (+0.3%) to 80.545%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
py/dynesty/sampler.py	7	8	87.5%
py/dynesty/bounding.py	13	15	86.67%
py/dynesty/sampling.py	5	7	71.43%
py/dynesty/dynamicsampler.py	57	61	93.44%
py/dynesty/utils.py	59	64	92.19%
<!--	Total:	153	167	91.62%	-->

Files with Coverage Reduction	New Missed Lines	%
py/dynesty/dynamicsampler.py	1	87.2%
py/dynesty/dynesty.py	1	80.65%
py/dynesty/utils.py	2	83.51%
<!--	Total:	4		-->

Totals
Change from base Build 969454074:	0.3%
Covered Lines:	3606
Relevant Lines:	4477

---

💛 - Coveralls

[segasai]

Hi, This branch is ready to be merged (it has a collection of various things that'd be nice to merge) I want to bring to particular attention this commit 54e268eeeb2d0e1551dbf2ff30d4020583d92199 as it fixes an annoying bug that I've introduced a couple of months ago in the update_interval calculation logic.

[joshspeagle]

How is this looking? Can I looking into merging this in? I also want to push the fix to https://github.com/joshspeagle/dynesty/commit/54e268eeeb2d0e1551dbf2ff30d4020583d92199 that we missed and get the updated version on pip sooner rather than later.

[segasai]

I think this is good to merge, with the only concern over the factor of two in front of the variance calculation (see another issue). If you are happy with that, I think this can be merged. (there is one more function where I'd like to reuse the compute_integrals(), but I can do it in a separate commit later)

[joshspeagle]

Okay, then let me look into resolving that first, we can add it in, and then merge.

[segasai]

Great! Thanks