Adding a `topn` parameter to the Exhaustive Feature Selector

[rasbt]

When running the ExhaustiveFeatureSelector, it is not unusual that an extraordinary large number of results get collected (due to the many combinations to evaluate). Some users reported that this leads to memory issues when calling the get_metric_dict() method that constructs a DataFrame containing the results.

Thus, it would be good to add a parameter like "topn" for the get_metric_dict() method that only returns the top n results.

Also, there seems to be an inconsistency that .subsets_ for efs strats at key 0 whereas sfs starts at 1 which should ideally be fixed.

This issue is based on the discussion on the mailinglist at https://groups.google.com/forum/#!topic/mlxtend/2h7WNBn_A38

[ScorpionXiezi]

I will be working on this issue. Follow up will be posted this week.

[rasbt]

Sounds good. Please note that I have to enter the grades on Monday, so please don't wait too long.

[sagnik1511]

Is this issue still alive? I'll be happy to contribute.

[rasbt]

Yes, it's still alive! And I would definitely welcome contributions! 🙌