Skip to content

Conversation

@alexfikl
Copy link
Collaborator

@alexfikl alexfikl commented Jan 6, 2025

This should fix the new failures. As far as I can tell, this is due to scipy 1.15 having a rewritten interp_decomp implementation.

>               assert tgt_error < ctol * case.id_eps
E               AssertionError: assert np.float64(7.672159794924751e-08) < (6 * 1e-08)
E                +  where 1e-08 = TorusTestCase(name='torus', ambient_dim=3, knl_class_or_helmholtz_k=0, knl_kwargs={}, bc_type='dirichlet', side=-1, qb...ighted_proxy=None, proxy_source_cluster_builder=None, proxy_target_cluster_builder=None, neighbor_cluster_builder=None).id_eps

Comment on lines +343 to +350
from scipy import __version__

kwargs = {}
if __version__ >= "1.15.0":
if rng is None:
rng = np.random.default_rng()

kwargs["rng"] = rng
Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Had to start using rng in here because scipy.linalg.interpolative.seed is a noop on the latest version, so the tests kept failing due to randomness.

@inducer inducer merged commit 94c40dc into inducer:main Jan 7, 2025
9 checks passed
@inducer
Copy link
Owner

inducer commented Jan 7, 2025

LGTM, thanks!

@alexfikl alexfikl deleted the fix-skel branch January 7, 2025 19:06
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants