Support (N, N) affines in rescale_affine

[SID-6921]

Summary

rescale_affine() documents its affine parameter as

(N, N) array-like
NxN transform matrix in homogeneous coordinates representing an affine transformation from an (N-1)-dimensional space to an (N-1)-dimensional space.

but the implementation slices affine[:3, :3], so only a 4x4 affine actually works. voxel_sizes(), which rescale_affine() calls, generalises correctly and returns N - 1 values, so the two disagree and the multiply raises:

>>> import numpy as np
>>> from nibabel.affines import rescale_affine
>>> aff = np.array([[2.0, 0.0, 10.0], [0.0, 3.0, 20.0], [0.0, 0.0, 1.0]])
>>> rescale_affine(aff, (64, 48), (1.0, 1.0))
ValueError: operands could not be broadcast together with shapes (3,3) (2,)

The same happens for a 5x5 affine.

Change

Take the dimensionality from the affine rather than assuming 3. apply_affine() and from_matvec() are already dimension-agnostic, so nothing else needed adjusting. The 4x4 path is unchanged.

np.asarray() is applied to affine first so that array-likes work, matching what the docstring promises.

Tests

Added test_rescale_affine_2d, which checks the returned affine has the right shape and zooms and that the documented invariant - the RAS location of the central voxel is preserved - still holds. It fails with a ValueError before the change.

pytest nibabel/tests/test_affines.py passes (8 passed).

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.48%. Comparing base (a5b68f4) to head (465bcc8).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1553   +/-   ##
=======================================
  Coverage   95.48%   95.48%           
=======================================
  Files         209      209           
  Lines       30045    30065   +20     
  Branches     4492     4492           
=======================================
+ Hits        28689    28709   +20     
  Misses        925      925           
  Partials      431      431           

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[effigies]

Thanks for this! A couple small comments, but this is definitely makes sense given that we mostly do write these without limiting to 3-dimensional data.

[effigies]

Thanks! Good to merge when tests pass.