Martin Stancsics

Can confirm it on my system as well. It looks to be an issue with altair_saver, not Altair itself. Also, most likely related to https://github.com/altair-viz/altair_saver/issues/70 It is supposedly fixed in...

Thanks for the reply! I totally agree that including it in the docs might be enough. It is an advanced feature after all, and users that need it can be...

In the meantime, here is a slightly more thought through shell function to achieve the same result: ```bash gwsl() { if [ "$#" -eq 0 ]; then echo "Set and...

AFAIK `formulaic` [provides support](https://matthewwardrop.github.io/formulaic/guides/contrasts/#guaranteeing-structural-full-rankness) for this with the `ensure_full_rank=False` and`spans_intercept=False` parameters.

I believe the reason for the nightly failure is [this deprecation](https://github.com/pandas-dev/pandas/pull/53043) came into effect in pandas 3.0.0dev0:

Glum has a built-in [method](https://glum.readthedocs.io/en/latest/glm.html#glum.GeneralizedLinearRegressor.wald_test) for performing a [Wald test](https://en.wikipedia.org/wiki/Wald_test), which (among others) can be used to test for the joint significance of a number of variables (e.g. the levels...

Agreed. While implementing confidence intervals for the predicted mean should be relatively straightforward, I'm not sure how useful they would be for the reasons @MatthiasSchmidtblaicherQC mentioned. For prediction tasks, people...

I agree, If we change `covariance_matrix` we should change `wald_test`, too. Putting required arguments after optionals looks weird at first, but it would be possible because they are "technically optional",...

> Should we change the default alpha in version 3 to `alpha=0`? There was a similar discussion for scikit-learn a few years ago: [article](https://ryxcommar.com/2019/08/30/scikit-learns-defaults-are-wrong/), [HN thread](https://news.ycombinator.com/item?id=21416702), There's also a bunch...

🎉