Sebastian Raschka

Hard to tell. If you are using scikit-learn decision trees, they set the threshold at x_i

Based on how complicated the function has become recently, this should really be a `plot_decision_regions_3d` to not further complicate things.

Unfortunately, I think you are right and this is currently not possible. The `category_scatter` function is among the oldest ones in mlxtend, and the API is not ideal. To address...

Hm, good question. The example in the documentation (http://rasbt.github.io/mlxtend/user_guide/evaluate/bias_variance_decomp/#example-3-tensorflowkeras-support) was ran with having TensorFlow v2.3.1 installed on my machine and only tried with `Sequential`. Maybe ```python if estimator.__class__.__name__ in ['Sequential',...

If it is not too much hassle/does not require big changes in the code, I would really welcome PRs that add Dask support.

Sorry, there's currently no option for that. A "`must_contain=('some item', 'some other item', ...)`" feature could be a useful enhancement though.

I agree with you @harenbergsd . In addition, while adding a check for whether frequent itemsets contain a particular item may look useful in practice, I am not sure if...

Yes, https://github.com/rasbt/mlxtend/pull/646 should address that.

Thanks a lot for the PR, this is very exciting! Big picture-wise, there are a few thoughts. 1) What do we do with the existing `ExhaustiveFeatureSelector` and `SequentialFeatureSelector`? We could...

> I think EFS is straightforward, but SFS I have to understand the "floating" logic a little better -- I gather this is just "search both forward and backward" but...