I would like to understand the difference between the KNN algorithm as implemented in scikit-learn (for example, using the kdtree algorithm)

https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsRegressor.html

and the one in scipy: https://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.KDTree.query.html

It's the same for me, but I'm not sure if I'm wrong or not ?

Thank you,

Asmae