AttributeError: 'LocalClassifierPerNode' object has no attribute '_validate_data'

[pbahr]

Describe the bug
Hi there,

I have pip installed the library and tried to run the Hello HiClass code.

Running the code, I get this error:

AttributeError: 'LocalClassifierPerNode' object has no attribute '_validate_data'

When I looked at the code, I see this line in the hiclass/HierarchicalClassifier.py file:

def _pre_fit(self, X, y, sample_weight):
# Check that X and y have correct shape
# and convert them to np.ndarray if need be

    if not self.bert:
        self.X_, self.y_ = self._validate_data(
            X, y, multi_output=True, accept_sparse="csr", allow_nd=True
        )

...

But I couldn't find the implementation for _validate_data() anywhere in the other classes, e.g. LocalClassifierPerNode

To Reproduce
Steps to reproduce the behavior:

1. pip install hiclass['xai','ray']
2. run the Hello HiClass code
3. You'll probably get the same error as me

Expected behavior
Print out the predictions for the test set

Desktop (please complete the following information):

• Running the code in the notebook environment for Azure Machine Learning
• hiClass 4.13.3
• numpy 1.21.6
• scipy 1.10.1
• scikit-learn 0.22.1

[pbahr]

Stack trace:
AttributeError Traceback (most recent call last)
Cell In[7], line 20
17 classifier = LocalClassifierPerNode(local_classifier=rf)
19 # Train local classifier per node
---> 20 classifier.fit(X_train, Y_train)
22 # Predict
23 predictions = classifier.predict(X_test)

File /anaconda/envs/azureml_py38/lib/python3.8/site-packages/hiclass/LocalClassifierPerNode.py:155, in LocalClassifierPerNode.fit(self, X, y, sample_weight)
134 """
135 Fit a local classifier per node.
136
(...)
152 Fitted estimator.
153 """
154 # Execute common methods necessary before fitting
--> 155 super().pre_fit(X, y, sample_weight)
157 # Initialize policy
158 self.binary_policy = self._initialize_binary_policy(calibration=False)

File /anaconda/envs/azureml_py38/lib/python3.8/site-packages/hiclass/HierarchicalClassifier.py:159, in HierarchicalClassifier.pre_fit(self, X, y, sample_weight)
154 def pre_fit(self, X, y, sample_weight):
155 # Check that X and y have correct shape
156 # and convert them to np.ndarray if need be
158 if not self.bert:
--> 159 self.X, self.y = self.validate_data(
160 X, y, multi_output=True, accept_sparse="csr", allow_nd=True
161 )
162 else:
163 self.X = np.array(X)

AttributeError: 'LocalClassifierPerNode' object has no attribute '_validate_data'