Jonathan Feinberg

I am not sure what you are asking here. If you want a response you need to clearify.

Right, you mean why not use the empirical mean directly. No they are not equivalent, though they both converge towards the true value as the number increase. Take a look...

I'm not able to replicate the bug on my system. Mind posting your current version of `numpoly` and `numpy` by doing `pip freeze`? Also if you have only tried to...

My best guess is that on your windows machine the machine precision is lower than on mine. I've made the assert more lenient now in version 4.3.5. Place check it...

Hm... the building of the docs depend on the existence of indexes for external docs to numpy, scipy etc. Looks like scipy reference guide is down and its index with...

Still challanging for me to diagnose. Could you perhaps try using `python -O` to disable the safety checks? Does it produce resultss similar to the Monte Carlo results?

That is an excellent idea. Once upon a time I did that for many of the distributions, but scipy did not have that particular distribution long ago and I have...

No, unfortunatly, this is out of scope for this software.

I am having a little bit of an issue accessing journal articles at the moment, so I can not answer concretely, but I still think I get what is going...

The good thing about sparse grid is that it scales a lot better in higher dimensions (when nestedness works as expected, which your example does not). However, it also has...