`DecodedFeatures.get()` does not works for `numpy>=2.0`

[ganow]

Behavior

Calling DecodedFeatures.get() under an environment installing numpy>=2.0 raises the following error:

RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/project-dir/.venv/lib/python3.11/site-packages/bdpy/dataform/features.py", line 30, in _load_array_with_key
    return sio.loadmat(path)[key]
           ^^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/scipy/io/matlab/_mio.py", line 234, in loadmat
    MR, _ = mat_reader_factory(f, **kwargs)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/scipy/io/matlab/_mio.py", line 80, in mat_reader_factory
    raise NotImplementedError('Please use HDF reader for matlab v7.3 '
NotImplementedError: Please use HDF reader for matlab v7.3 files, e.g. h5py

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/python-share-dir/lib/python3.11/multiprocessing/pool.py", line 125, in worker
    result = (True, func(*args, **kwds))
                    ^^^^^^^^^^^^^^^^^^^
  File "/python-share-dir/lib/python3.11/multiprocessing/pool.py", line 48, in mapstar
    return list(map(*args))
           ^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/bdpy/dataform/features.py", line 33, in _load_array_with_key
    return hdf5storage.loadmat(path)[key]
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/hdf5storage/__init__.py", line 1764, in loadmat
    options = Options(marshaller_collection=marshaller_collection)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/hdf5storage/__init__.py", line 260, in __init__
    self.marshaller_collection = MarshallerCollection()
                                 ^^^^^^^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/hdf5storage/__init__.py", line 902, in __init__
    self._builtin_marshallers = [m() for key, m in dict(
                                ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/hdf5storage/__init__.py", line 902, in <listcomp>
    self._builtin_marshallers = [m() for key, m in dict(
                                 ^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/hdf5storage/Marshallers.py", line 492, in __init__
    np.bytes_, np.unicode_, np.object_]
               ^^^^^^^^^^^
  File "/project-dir/.venv/lib/python3.11/site-packages/numpy/__init__.py", line 400, in __getattr__
    raise AttributeError(
AttributeError: `np.unicode_` was removed in the NumPy 2.0 release. Use `np.str_` instead.
"""

The above exception was the direct cause of the following exception:

AttributeError                            Traceback (most recent call last)
Cell In[12], line 1
----> 1 decoded_features = decoded_features_dataset.get(layer=layer_path_name, subject=subject, roi=roi)

File /project-dir/src/dists_like_loss/dataset.py:112, in DecodedFeaturesDataset.get(self, layer, subject, roi, fold, label)
    111 def get(self, layer=None, subject=None, roi=None, fold=None, label=None):
--> 112     return self._decoded_features_store.get(
    113         layer=layer,
    114         subject=subject,
    115         roi=roi,
    116         fold=fold,
    117         label=label,
    118     )

File /project-dir/.venv/lib/python3.11/site-packages/bdpy/dataform/features.py:390, in DecodedFeatures.get(self, layer, subject, roi, fold, label, image)
    388 else:
    389     with Pool(processes=num_parallel) as pool:
--> 390         features = np.concatenate(pool.map(partial(_load_array_with_key, self.__file_key), files), axis=0)
    392 if self.__squeeze:
    393     features = np.squeeze(features)

File /python-share-dir/lib/python3.11/multiprocessing/pool.py:367, in Pool.map(self, func, iterable, chunksize)
    362 def map(self, func, iterable, chunksize=None):
    363     '''
    364     Apply `func` to each element in `iterable`, collecting the results
    365     in a list that is returned.
    366     '''
--> 367     return self._map_async(func, iterable, mapstar, chunksize).get()

File /python-share-dir/lib/python3.11/multiprocessing/pool.py:774, in ApplyResult.get(self, timeout)
    772     return self._value
    773 else:
--> 774     raise self._value

AttributeError: `np.unicode_` was removed in the NumPy 2.0 release. Use `np.str_` instead.

Idea on the solution

This problem is probably caused by using hdf5storage.loadmat. In fact, if I run hdf5storage.loadmat() in an environment with numpy>=2.0, I get the same error. The last commit on the hdf5storage GitHub repository was two years ago, so we cannot expect continuous development.
Based on this observation, I propose to change the following loader function:

bdpy/bdpy/dataform/features.py

Lines 25 to 32 in d86199f

	import hdf5storage
	def _load_array_with_key(key: str, path: str) -> np.ndarray:
	try:
	return sio.loadmat(path)[key]
	except (NotImplementedError, ValueError):
	return hdf5storage.loadmat(path)[key]

as

import h5py

def _load_array_with_key(key: str, path: str) -> np.ndarray:
    try:
        return sio.loadmat(path)[key]
    except (NotImplementedError, ValueError):
        with h5py.File(path, "r") as f:
            return f[key][:].transpose()

hdf5storage is also used in other places, and I think it is necessary to gradually replace it with h5py.

$ grep -r hdf5storage . --include "*.py"
./bdpy/dataform/datastore.py:import hdf5storage
./bdpy/dataform/datastore.py:        r = hdf5storage.loadmat(fpath)[self.__variable]
./bdpy/dataform/sparse.py:import hdf5storage
./bdpy/dataform/sparse.py:            return hdf5storage.loadmat(fname)[key]
./bdpy/dataform/sparse.py:        hdf5storage.savemat(fname,
./bdpy/dataform/sparse.py:    hdf5storage.savemat(fname,
./bdpy/dataform/sparse.py:        hdf5storage.savemat(fname, {key: {u'__bdpy_sparse_arrray': True,
./bdpy/dataform/sparse.py:        data = hdf5storage.loadmat(fname)[key]
./bdpy/dataform/features.py:import hdf5storage
./bdpy/dataform/features.py:        return hdf5storage.loadmat(path)[key]
./bdpy/dataform/features.py:            self.__feat_index_table = hdf5storage.loadmat(feature_index)['index']
./bdpy/dataform/features.py:    hdf5storage.savemat(save_file, {'feat': feature})
./tests/dataform/test_features.py:import hdf5storage
./tests/dataform/test_features.py:            hdf5storage.savemat(