Problem if scipy version is 1.13.0

[satabol]

Hello. I'm user of your library. It is cool! But I have some trouble after today update. I find out that if scipy=1.13.0 then result is not good. If force reinstall to scipy=1.12.0 then result is good. Step to reproduce:

pip.exe install pyacvd

pip.exe install --force-reinstall scipy==1.12.0

code:

import pyacvd
from pyvista import PolyData
verts_i=[(1.0, 1.0, 1.0), (1.0, 1.0, -1.0), (1.0, -1.0, 1.0), (1.0, -1.0, -1.0), (-1.0, 1.0, 1.0), (-1.0, 1.0, -1.0), (-1.0, -1.0, 1.0), (-1.0, -1.0, -1.0)]
faces_i=[[0, 4, 6, 2], [3, 2, 6, 7], [7, 6, 4, 5], [5, 1, 3, 7], [1, 0, 2, 3], [5, 4, 0, 1]]
pdmesh = PolyData.from_regular_faces( verts_i, faces_i)
clust = pyacvd.Clustering(pdmesh)
clust.cluster(1000, maxiter=100)
remesh = clust.create_mesh()
remesh.regular_faces

Windows 11, Python 3.10.11

[akaszynski]

Can reproduce. Expect a fix by the end of the week.

[satabol]

Thank you! It is work now! Here is a first result:

But I install this version from source code. Can you say when one can install this version from pip repository?

P.S. I'm one of developers of Sverchok plugin for Blender. Here is a docs of node of your pyacvd library.

[akaszynski]

This is awesome! Nice to see pyacvd and pyvista used within a plugin for blender.

Right now we're delayed due to an account issue. We're tracking it at pypi/support #4016, but we might be able to have it resolved next Monday. I'll keep you informed, but expect a new release by the end of next week.

[satabol]

@akaszynski Thank you!

[nicolascedilnik]

I confirm that reverting scipy to 1.12 worked for me. Thanks @akaszynski for this workaround!

[akaszynski]

@satabol, this fix is now in pyacvd v0.2.11.

[satabol]

Thank you! I test it. All work fine!