0.1.0-beta Release

[jessdtate]

The beta release for UncertainSCI is here. This version includes some bug fixes from the alpha release, a new feature to adaptively find the optimal order of the problem, and some new examples to try.

---

This discussion was created from the release 0.1.0-beta Release.

[ww382]

Hi!

I downloaded the beta version of the source code and installed UncertainSCI following the instructions here.

When I proceeded to run demo/build_pce.py. I received an error message indicating

Traceback (most recent call last):
File "../demos/build_pce.py", line 10, in
from UncertainSCI.distributions import BetaDistribution
ModuleNotFoundError: No module named 'UncertainSCI'

I would appreciate if you can provide me suggestions on how to resolve this error.

For your reference, I listed the packages that were installed in my conda environment below.

Package Version
certifi 2020.4.5.1
cycler 0.10.0
kiwisolver 1.2.0
matplotlib 3.1.3
numpy 1.17.2
packaging 20.3
pip 21.0.1
pyparsing 2.4.7
python-dateutil 2.8.1
scipy 1.4.1
setuptools 58.0.4
six 1.15.0
UncertainSCI 0.0.0
wheel 0.37.0

[jessdtate]

Are you running on big sur, or what operating system?

[ww382]

yes, I'm running on big sur version 11.5.2.

[jessdtate]

There is a know bug with matplotib and big sur. You can comment out the plotting function and the script will finish. I'll check the status on that bug and see when it might be fixed

[jessdtate]

Looks like there might be a fix by uninstalling and reinstalling matplotlib.

https://code2care.org/python/python-matplotib-segmentation-flult-11-macos-big-sur

[ww382]

uninstalling and reinstalling matplotlib works! Thank you so much!

[ww382]

I've been running the demo scripts and trying to understand how UncertainSCI works. How do I output the Sobol sensitivity indices, Np, and Ns as shown in, for example, Table A1 in Burk et al?

Burk, KM, Narayan, A, Orr, JA. Efficient sampling for polynomial chaos-based uncertainty quantification and sensitivity analysis using weighted approximate Fekete points. Int J Numer Meth Biomed Engng. 2020; 36:cnm03395.