[BUG] pip install graphistry[umap_learn] finds cupy import error when running umap even if engine=umap_learn

[silkspace]

Describe the bug
pip install graphistry[umap_learn] finds cupy import error when running umap even if engine=umap_learn. Pip installing cupy doesn't resolve problem.

To Reproduce

g.umap(engine='umap_learn')

Expected behavior
Confirmed that in google.collab this works, but it did not when client tried.

Actual behavior
Error:

    from cupy import _core  # NOQA
  File “/usr/local/lib/python3.10/dist-packages/cupy/_core/__init__.py”, line 3, in <module>
    from cupy._core import core  # NOQA
ImportError: libcuda.so.1: cannot open shared object file: No such file or directory
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
  File “/usr/local/lib/python3.10/dist-packages/graphistry/utils/lazy_import.py”, line 11, in lazy_cudf_import
    import cudf  # type: ignore
  File “/usr/local/lib/python3.10/dist-packages/cudf/__init__.py”, line 12, in <module>
    import cupy
  File “/usr/local/lib/python3.10/dist-packages/cupy/__init__.py”, line 19, in <module>
    raise ImportError(f’’'
ImportError:
================================================================
Failed to import CuPy.

Browser environment (please complete the following information):

• OS: [Windows]
• Browser [chrome]
• version []

Graphistry GPU server environment

• Where run [Google Collab]

PyGraphistry API client environment

• Where run [Collab Graphistry 2.35.9 Jupyter]
• Version [0.34.3]
• Python Version [e.g. Python 3.7.7]

[lmeyerov]

Are we confident they pip install graphistry[umap-learn] and not pip install graphistry ?

Related: The warning should get friendlier w/ #578 (comment) , on my TODOs to test a bit more & land