Change in precision of stable entropy bins with NumPy>=1.25.0

[rhugonnet]

Identified in #160. Tracked down that the change happens since NumPy 1.25.0 (CI fails for above, but not any of the 1.24.X or older), but we currently have no idea which function exactly triggers it... As only Python 3.9/3.10 use these most recent NumPy versions in CI, this is where tests fail with:

 =========================== short test summary info ============================
FAILED skgstat/tests/test_variogram.py::TestVariogramArguments::test_binning_method_stable - AssertionError: 
Arrays are not almost equal to 1 decimals

Mismatched elements: 5 / 10 (50%)
Max absolute difference: 0.21285498
Max relative difference: 0.03980185
 x: array([ 4.1,  8.6, 12.8, 17.2, 21.6, 25.2, 29.9, 33.1, 38.7, 43. ])
 y: array([ 4.3,  8.4, 12.8, 17.1, 21.4, 25.2, 29.9, 33.2, 38.5, 42.8])
FAILED skgstat/tests/test_variogram.py::TestVariogramArguments::test_binning_method_stable_fix_bins - AssertionError: 
Arrays are not almost equal to 1 decimals

Mismatched elements: 3 / 10 (30%)
Max absolute difference: 0.49334201
Max relative difference: 0.01147307
 x: array([ 4.2,  8.5, 12.8, 17.1, 21.3, 25.5, 29.6, 33.4, 37.4, 43.5])
 y: array([ 4.2,  8.6, 12.8, 17.1, 21.2, 25.5, 29.3, 33.2, 37.4, 43. ])
FAILED skgstat/tests/test_variogram.py::TestVariogramArguments::test_binning_method_stable_maxiter - AssertionError: 
Arrays are not almost equal to 1 decimals

Mismatched elements: 5 / 10 (50%)
Max absolute difference: 0.21285498
Max relative difference: 0.03980185
 x: array([ 4.1,  8.6, 12.8, 17.2, 21.6, 25.2, 29.9, 33.1, 38.7, 43. ])
 y: array([ 4.3,  8.4, 12.8, 17.1, 21.4, 25.2, 29.9, 33.2, 38.5, 42.8])

For now, the tests are temporarily fixed by changing test precision from 1 to 0 decimals.