Jonathan Feinberg

I don't know. You mind making a minimal code example that illustrate how you used chaospy to estimate the standard deviations?

I can't be sure, since I don't have the full picture, but I see a potential problem with `h` is almost 100 times as big as `rho`. I highly recommend...

Not all problems are a good fit for PCE. The fact the your expansions converges quickly with samples, but not with order, might indicate that your problem is an example...

PCE assumes that the `samples` comes from a probability distribution. Irrespective of technique, it is needed for e.g. polynomial generation. Assuming you have that, you can follow the [quick tutorial](https://chaospy.readthedocs.io/en/master/user_guide/quick_tutorial.html)...

I just updated `numpoly`. Update to 1.2.13 and this should be resolved. Let me know if it doesn't solve the issue.

This should be fixed. I just forgot to close ithe issue. Can you ensure that the following packages are up to date: pip install -U chaospy numpoly If this doesn't...

Jupp. I missed one. Update `numpoly` to version 1.2.14 and you should be good to go.