The presence of this method causes the numpy floatmul to use it in numpy 2.1 which results in the units not being tagged as expected and the example failing.

PR summary

PR checklist

"closes #0000" is in the body of the PR description tolink the related issue
new and changed code istested
Plotting related features are demonstrated in anexample
New Features andAPI Changes are noted with adirective and release note
Documentation complies withgeneral anddocstring guidelines

FIX: remove __array__ method on BasicUnit

2b3d142

The presence of this method causes the numpy float __mul__ to use itin numpy 2.1 which results in the units not being tagged as expected andthe example failing.

tacaswell added this to thev3.9-doc milestone

Aug 21, 2024

github-actionsbot added the Documentation: examplesfiles in galleries/examples label

Aug 21, 2024

Copy link

MemberAuthor

tacaswell commentedAug 21, 2024

Sorry, I did this a while ago but forgot to actually open the PR!

Remove __array__ from BasicUnit

b4ab05d

timhoffm approved these changes

Aug 21, 2024

View reviewed changes

QuLogic requested changes

Aug 21, 2024

View reviewed changes

Copy link

Member

QuLogic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others.Learn more.

This breaks the docs with NumPy 2.0.1.

Update missing references for moved lines

108565a

Copy link

Member

QuLogic commentedAug 21, 2024•
edited
Loading

So the bisect points this tonumpy/numpy@682cd04 vianumpy/numpy#26904.

Reading through the array API docs, it seems like setting__array_priority__ higher (default is 0) may also work, in order to ensure that our__array_wrap__ method is called regardless of version, and we don't end up with a bare array. Alternatively, we could implement__array_ufunc and/or__array_function__.