Stefan van der Walt

Understood about the mixing. What I meant was that currently you have to do something different for NumPy arrays than for Array-API-compatible arrays; I just want to see that gap...

/cc @grlee77 @hmaarrfk Any lessons from [the skimage fetcher](https://github.com/scikit-image/scikit-image/blob/main/skimage/data/_fetchers.py) that should apply here?

Thanks, @nrweir, looking forward to your first PR!

Compilation with OpenMP see also https://github.com/scikit-image/scikit-image/pull/1570/files

Curious why we convert to float internally, given that there are comparisons involved.

Why not just slice the image before computing regionprops?

@ldes89150 bug confirmed. @jni agreed.

On Thu, 11 Apr 2019 16:42:53 -0700, Mark Harfouche wrote: > Not sure the best way to deal with this that won't completely kill > performance. It's okay to have...

Do you think something like `dask.delayed` could fulfill a similar function? That way, a pipeline can be defined as a Python function, which should be a bit more readable. Maybe...

For now, I think the best way to chain operations is to use Python. Unless you have very well defined APIs for functions in the pipeline (e.g. sklearn classifieds), a...