Roman Feldbauer

You could try inheriting from the [`ClassifierMixin`](https://github.com/scikit-learn/scikit-learn/blob/e113897235feaf309eaaed24001ca96f3608602f/sklearn/base.py#L635) which provides the `score()` method. Or explicitly write the method. If you take a look at the linked code, it's literally two lines...

Please post the error message you get.

I quickly googled these error messages and some reports on insufficient memory popped up. But this is a wild guess. I fear I cannot give more support on falconn compilation...

This is addressed by #45

Puffinn tests are still disabled on macos, because it appears to be less precise.

Thanks for reporting this. Allowing all spaces that nmslib supports is certainly desirable.

Fractional norms and various other metrics supported by `nmslib` will be available in the upcoming scikit-hubness v0.30. (No PyPI yet, but one can already install from the sources).

Quickstart example of hubness estimation, hubness reduction and usage in sklearn workflows updated in the readme in #105. Userguide and further examples upcoming.

With the removal of `falconn` from scikit-hubness, there is no radius search left. While nice to have, radius search is not of particularly high interest to hubness research. Therefore, removed...

I've been looking into this for my [scikit-hubness](https://github.com/VarIr/scikit-hubness). There's a couple of alternatives for which I found ways to serialize indexes: - [Puffinn](https://github.com/puffinn/puffinn) also provides LSH - [NGT](https://github.com/yahoojapan/NGT) - [Annoy](https://github.com/spotify/annoy)...