The GUDHI library is a generic open source C++ library, with a Python interface, for Topological Data Analysis (TDA) and Higher Dimensional Geometry Understanding.
While testing a new compilation environment with up-to-date third parties libraries, I faced errors with perslay tests:
File "/workspace/gudhi/gudhi-devel/build/src/python/gudhi/tensorflow/perslay.py", line 132, in __init__
super().__init__(dynamic=True, **kwargs)
File "/miniconda3/envs/custom-cgal/lib/python3.10/site-packages/keras/src/layers/layer.py", line 263, in __init__
raise ValueError(
ValueError: Unrecognized keyword arguments passed to GaussianPerslayPhi: {'dynamic': True}
The perslay python tests is ok with tensorflow 2.11.0, but fails with tensorflow 2.15.1
Edit: I was pretty sure the dynamic was removed with tensorflow 2.15.1 (that seems strange), but after testing master branch again with python 3.11, it seems to occur on tensorflow version 2.16.
While testing a new compilation environment with up-to-date third parties libraries, I faced errors with perslay tests:
The perslay python tests is ok with tensorflow 2.11.0, but fails with tensorflow 2.15.1