Christian Lorentzen

I'm a user not involved in neither dplyr nor arrow and the current state of names, for me, is confusing. Maybe we should trust @jorgecarleitao with this, as he is...

> To the extent that [scikit-learn](https://github.com/scikit-learn/scikit-learn/) or [glum](https://github.com/Quantco/glum) could potentially continue supporting the sparse structures, that would definitely be a win. Both support sparse feature matrices and will continue to...

> Not trying to block this PR but wouldn't it make sense in the long run to implement `logsumexp` in cython or C so that it can be exposed via...

Is the naming already decided? While pytorch calls it `softplus`, it‘s not a common name that I have met often. Something like `log1pexp` is also a bit cryptic. I just...

Tensorflow also calls it softplus. That name seems to stem from [Dugas et al 2001](https://papers.nips.cc/paper_files/paper/2000/hash/44968aece94f667e4095002d140b5896-Abstract.html). In that case, better keep the name `softplus`.

In my experience, quantiles are very important in practice and occur all over the places. > > Which ones? Could we get away with only a default method, and hence...

> adding an implementation to SciPy in terms of array API functions @mdhaber You might want to contact/coordinate with scikit-learn for that. (Apart from numpy 2.0 not yet being the...

@lucascolley Out of interest: Who decided when that this is *not planned*?

But filenames translate to python (sub) modules and I'd like to filter those and couldn't get it to work as needed. Instead, `gen_ref_pages.py` has ```python set(Path("src").rglob("*.py")) - set(Path("src").rglob("*tests/*.py")) - set(Path("src").rglob("_*/*.py"))...

I'm thinking of a new function `compute_score` as the `compute_bias`. It's just simple t-tests, see the code of compute_bias.