Avoid unit test failures with newer sklearn

[mgalardini]

With sklearn > 1.5.0 I encountered we now get a different non-metric MDS

[mgalardini]

I found a way to have the same environment as github actions locally, because I am not seeing these errors in my setup:

| =========================== short test summary info ============================                                                                                                                                 
| FAILED tests/enet_test.py::TestFitEnet::test_fit_binary - IndexError: index 1 is out of bounds for axis 1 with size 1                                                                                            
| FAILED tests/enet_test.py::TestFitEnet::test_fit_continuous - AttributeError: module 'numpy' has no attribute 'NAN'                                                                                              
| ================== 2 failed, 42 passed, 54 warnings in 5.25s ===================

Hopefully I'll find a way to fix it, will push commit here