Support numpy >= 2.4 with backward-compatible fixes#5
Merged
Conversation
numpy 2.4 expired the ndim>0-to-scalar deprecation: np.float64(arr) now returns an array for 1-element arrays instead of collapsing to a scalar, so PyACECalculator's results['energy'] leaked out as a shape-(1,) array from get_potential_energy(). Use float() on the 0-d energy values instead, which behaves identically on all numpy versions, and astype(np.float64) for per-atom arrays (also fixing energies_dev degrading to a scalar for single-atom structures on older numpy). numpy 2.4 also removed np.trapz (deprecated alias of np.trapezoid since 2.0); pyace.radial now picks whichever exists. Verified by building the package and running tests/test_PyACECalculator.py against numpy 1.26.4, 2.3.5 and 2.4.6. Relax the numpy requirement to <3 accordingly. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015Ui6Drmop9gZXo1fgn8yVg
pmrv
commented
Jul 16, 2026
pmrv
commented
Jul 16, 2026
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015Ui6Drmop9gZXo1fgn8yVg
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Pull Request Template
Description of Changes
Follow-up to #3: makes the code compatible with numpy ≥ 2.4 in a backward-compatible way and relaxes the requirement from
numpy<2.4tonumpy<3.Two numpy 2.4 breakages are fixed, both working identically on old and new numpy:
src/pyace/asecalc.py). numpy 2.4 expired the "conversion of an array with ndim > 0 to a scalar" deprecation, sonp.float64(energy.reshape(-1,))started returning a shape-(1,)array instead of a scalar, andget_potential_energy()leaked an array of the total energy. The results dicts now usefloat()on the 0-d/scalar values (energy,free_energy, and the ensembleenergy_std/energy_dev), which behaves identically on every numpy version, and.astype(np.float64)on per-atom arrays (energies_dev,forces_dev) — this also fixes those arrays silently degrading to scalars for single-atom structures on older numpy.np.trapzremoval (src/pyace/radial.py).np.trapzwas a deprecated alias ofnp.trapezoidfrom numpy 2.0 and was removed in 2.4. The module now picksnp.trapezoidwhen it exists and falls back tonp.trapzon numpy 1.x.A regression assertion is added to
tests/test_PyACECalculator.pyso the energy coming back as an array would fail the tests.Verification: built the package and ran
tests/test_PyACECalculator.py(13 tests) against numpy 1.26.4 (exercises thenp.trapzfallback), 2.3.5, and 2.4.6 — all pass on each. CI on this PR additionally covers the full test suite, which will now resolve numpy 2.4.x.Checklist
License Agreement
By submitting this pull request, I agree that:
🤖 Generated with Claude Code
https://claude.ai/code/session_015Ui6Drmop9gZXo1fgn8yVg
Generated by Claude Code