Optional packages torch-sparse, torch-cluster, and pyg-lib are missing dependencies

[ErasonElisabeth]

😵 Describe the installation problem

Installation of torch-sparse, torch-cluster and pyg-lib results in missing or unresolved dependencies. I repeated the installation process multiple times today, each time with a clean venv, tried different installation orders, checked out to see if using cuda-versions makes a difference (doesn't), went back to older versions of torch, no difference either. (See below under environment for precise installation script).
I keep running into the same problem when trying to import torch_geometric afterwards:

C:\Users\...\Lib\site-packages\torch_geometric\typing.py:68: UserWarning: An issue occurred while importing 'pyg-lib'. Disabling its usage. Stacktrace: Could not find module 'C:\Users\...\Lib\site-packages\libpyg.pyd' (or one of its dependencies). Try using the full path with constructor syntax.
  warnings.warn(f"An issue occurred while importing 'pyg-lib'. "
C:\Users\...\Lib\site-packages\torch_geometric\typing.py:97: UserWarning: An issue occurred while importing 'torch-cluster'. Disabling its usage. Stacktrace: Could not find module 'C:\Users\...\Lib\site-packages\torch_cluster\_fps_cpu.pyd' (or one of its dependencies). Try using the full path with constructor syntax.
  warnings.warn(f"An issue occurred while importing 'torch-cluster'. "
C:\Users\...\Lib\site-packages\torch_geometric\typing.py:124: UserWarning: An issue occurred while importing 'torch-sparse'. Disabling its usage. Stacktrace: Could not find module 'C:\Users\...\Lib\site-packages\torch_sparse\_convert_cpu.pyd' (or one of its dependencies). Try using the full path with constructor syntax.
  warnings.warn(f"An issue occurred while importing 'torch-sparse'.

The cited files (libpyg.pyd, _fps_cpu.pyd, _convert_cpu.pyd) are present at their supposed locations, so I am assuming a dependency hidden within is not correctly resolved.
This is my complete list of installed packages:

aiohappyeyeballs   2.4.3
aiohttp            3.11.7
aiosignal          1.3.1
attrs              24.2.0
certifi            2024.8.30
charset-normalizer 3.4.0
colorama           0.4.6
filelock           3.16.1
frozenlist         1.5.0
fsspec             2024.10.0
idna               3.10
Jinja2             3.1.4
MarkupSafe         3.0.2
mpmath             1.3.0
multidict          6.1.0
networkx           3.4.2
numpy              2.1.3
pip                24.2
propcache          0.2.0
psutil             6.1.0
pyg_lib            0.4.0+pt25cpu
pyparsing          3.2.0
requests           2.32.3
scipy              1.14.1
setuptools         75.6.0
sympy              1.13.1
torch              2.5.1
torch_cluster      1.6.3+pt25cpu
torch-geometric    2.6.1
torch_scatter      2.1.2+pt25cpu
torch_sparse       0.6.18+pt25cpu
torch_spline_conv  1.2.2+pt25cpu
tqdm               4.67.1
typing_extensions  4.12.2
urllib3            2.2.3
yarl               1.18.0

Environment

• PyG version: 2.6.1
• PyTorch version: 2.5.1
• OS: Win11
• Python version: 3.1.2
• CUDA/cuDNN version: none (cpu only)
• How you installed PyTorch and PyG (conda, pip, source):
  • clean python3.1.2 venv

     pip install torch
     pip install torch_geometric
     pip install pyg_lib torch_scatter torch_sparse torch_cluster torch_spline_conv -f https://data.pyg.org/whl/torch-2.5.1+cpu.html 
  

• Any other relevant information (e.g., version of torch-scatter):
  • torch_sparse 0.6.18+pt25cpu
  • torch_cluster 1.6.3+pt25cpu
  • pyg_lib 0.4.0+pt25cpu