Fix vector test for NumPy 2.4 - #5161

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 86.39%. Comparing base (f4d8ae4) to head (bfd3be3).
⚠️ Report is 7 commits behind head on master.

@@            Coverage Diff             @@
##           master    #5161      +/-   ##
==========================================
+ Coverage   85.51%   86.39%   +0.87%     
==========================================
  Files         282      282              
  Lines       59480    59480              
==========================================
+ Hits        50867    51385     +518     
+ Misses       8613     8095     -518     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

I am not familiar with this code. Just wanted to point out that np.array([[2.0], [0.0], [2.0], [1.0]]) and np.array([2.0, 0.0, 2.0, 1.0]) are not exactly the same. The first is a 4x1 two-dimensional array, the latter is a one-dimensional array with 4 elements.:

>>> np.array([[2.0], [0.0], [2.0], [1.0]])
array([[2.],
       [0.],
       [2.],
       [1.]])
>>> np.array([2.0, 0.0, 2.0, 1.0])
array([2., 0., 2., 1.])
>>> 

If that is intended, it is fine with me.

Exactly - but why would you want to do that here, and should it be supported?

i.e. Was the test doing something weird (in which case this fixes the test), or something reasonable that should still work with NumPy 2.4 (in which case we do need a code change, perhaps more like #5138).

Applying the change to the test now so that we can get the default branch building and passing cleanly, and do the release.

There may still be an underlyning issue that we will have to return too...