Jonathan Feinberg

The first 4 are documentation errors. I can fix it for the next release. Warnings isn't very high on my priorities. They are mostly docs and numpy operations that get...

Sure, updated now in version 3.3.7. But version 1.0.5 exists in pip, so I am unsure why this is an issue.

`cp.Uniform(5518.135, 5408.865)` is illigal. Try swapping the order of the arguments. I'll add checks for this in the next release.

See this post: https://stats.stackexchange.com/questions/70930/difference-between-sobol-indices-and-total-sobol-indices Is it possible that you have no higher order interaction between your components?

I right, I've seen that before. Your covariance matrix has a too large condition number. See #234 (9 posts down) for discussion and solution. I'll keep this issue open to...

Following the solution in the thread. As you are using Uniform not Normal, use `cp.J(cp.Uniform(-1, 1), cp.Uniform(-1, 1)` as your proxy distribution.

I am not sure if I understand your question, so please correct me if I am off base here. As you are getting the shape of your evaluations as 6x1,...

Okay, then I misunderstood. To understand what is going on, I think I need a little more info here. Typically in PCE we split up the problem domain into two...

Okay, that helps a lot. I wrote: evals = [my_model(x) for x in abscissas.T] Can I assume that in your case `my_model` returns a vector of length 6000? Your code...

So there are a few of things to note. When it comes to sum smaller or larger than one, your results seems to be valid. Sum of main indices can...