Update for Ax 1.0.0

[shuds13]

Ensure Ax examples work with Ax 1.0.0 with new unified API.

[shuds13]

Note that AxClientGenerator with Ax 0.5.0 does not work with numpy 2.4+ or pandas 3+

with numpy 2.4

[21](https://github.com/optimas-org/optimas/actions/runs/22084124691/job/63815051865#step:5:222)
  File "/usr/share/miniconda/envs/testing/lib/python3.12/site-packages/ax/modelbridge/transforms/winsorize.py", line 393, in _get_auto_winsorization_cutoffs_outcome_constraint
    q1 = np.percentile(Y, q=25, interpolation="lower")
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: percentile() got an unexpected keyword argument 'interpolation'

https://github.com/optimas-org/optimas/actions/runs/22084124691/job/63815051865

with pandas 3

In test_ax_generators.py/test_ax_client
  File "/home/shudson/miniconda3/envs/optimas_vocs/lib/python3.12/site-packages/ax/modelbridge/transforms/base.py", line 142, in transform_observations
    obs_data = self._transform_observation_data(observation_data=obs_data)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/shudson/miniconda3/envs/optimas_vocs/lib/python3.12/site-packages/ax/modelbridge/transforms/winsorize.py", line 167, in _transform_observation_data
    obsd.means[idx] = max(obsd.means[idx], self.cutoffs[metric_name][0])
    ~~~~~~~~~~^^^^^
ValueError: assignment destination is read-only

see
https://github.com/optimas-org/optimas/actions/runs/22508759859/job/65213177254