add docstring for learning

[senwu]

No description provided.

[lukehsiao]

There are also still many utilities that do not have documentation. If they are not meant to be public, they should be made private. Otherwise, they should be documented.

For example:

fonduer.learning.utils.print_scores(ntp, nfp, ntn, nfn, title='Scores')

fonduer.learning.utils.binary_scores_from_counts(ntp, nfp, ntn, nfn)

class fonduer.learning.utils.Scorer(test_candidates, test_labels, gold_candidate_set=None)

class fonduer.learning.utils.ModelTester(model_class, model_class_params, params_queue, scores_queue, X_train, X_valid, Y_valid, Y_train=None, b=0.5, beta=1, set_unlabeled_as_neg=True, save_dir='checkpoints', eval_batch_size=None)

All of these init parameters should be documented as well.

[lukehsiao]

Basically, Core Learning Objects are pretty much ready, just want to eliminate kwargs wherever possible.

The Learning Utilities are still missing several docstrings on initilizations for classes and some function parameters.

[lukehsiao]

The other issue we still need to address is documenting the wrappers for MeTal (LabelAnalyzer and LabelLearner) so that users know what functionality they provide.

Because MeTaL lacks documentation, we may need to do this ourselves for now.

[senwu]

Fixed! Merge to docstrings branch if LGTY. Thanks.