Scikit-learn >1.2 do not support the use of the Boston dataset from sklearn.datasets. This problem was raised back in december: https://github.com/Ekeany/Boruta-Shap/issues/111 and while one could make workaround importing Boston dataset from other sources, I am not an official maintainer. I merely suggest to replace sklearn.datasets toy dataset: load_boston() with load_diabetes().
What does this PR do?
This PR replaces the use of the load_boston() dataset in BorutaShap with load_diabetes() to avoid compatibility issues with scikit-learn >1.2.
Scikit-learn >1.2 do not support the use of the Boston dataset from sklearn.datasets. This problem was raised back in december: https://github.com/Ekeany/Boruta-Shap/issues/111 and while one could make workaround importing Boston dataset from other sources, I am not an official maintainer. I merely suggest to replace sklearn.datasets toy dataset: load_boston() with load_diabetes().
What does this PR do?
This PR replaces the use of the load_boston() dataset in BorutaShap with load_diabetes() to avoid compatibility issues with scikit-learn >1.2.
References
-Issue raised in December 2022: https://github.com/Ekeany/Boruta-Shap/issues/111 -Documentation for load_diabetes(): https://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_diabetes.html#sklearn.datasets.load_diabetes
Testing performed
I have tested the loading functionality, replacing load_diabetes() with load_boston(), on my local machine and the code runs without errors.
Known issues