simaki

Thanks a lot for a great suggestion @markdregan ! I haven't thought of that and think that would be a great feature to make fracdiff more convenient and sklearn compatible....

Another motivation: faster implementation of unit-root test (cf. https://github.com/fracdiff/fracdiff/issues/18#issuecomment-834430506)

Thank you for sharing a cool paper. The feature this issue addresses would allow users to make use of user-defined unit-root testers. Although we are not planning to implement any...

> When fit() is called, check and store column names in self.column_names > When get_feature_names() is called, return self.column_names if available I guess inheriting _OneToOneFeatureMixin and EstimatorMixin may grant our...

> I've tested the above and include test cases below. Very much appreciate it :) glad if you could add it to tests/ so we can see its victory via...

> I notice in my testing that the transformed array output shape is N-1 along axis=0. Is this by design? no, it should be N as you point out. Thank...

> I've not written tests before (but keen to learn). Given this new approach (inheriting _OneToOneFeatureMixin), what types of tests would be appropriate? Any pointers to similar ones existing? Your...

fyi https://scikit-learn.org/stable/auto_examples/release_highlights/plot_release_highlights_1_1_0.html#get-feature-names-out-available-in-all-transformers

Hi. #63 solves it. Many thanks.