Discriminative-lexicon-model

* Fixes a bug in build.

**Full Changelog**: https://github.com/quantling/discriminative_lexicon_model/compare/v2.0.0...v2.0.1

What's Changed
* Refactors functions in performance.py
* accuracy has now keyword-only arguments.
* predict_df has now keyword-only arguments.
* `accuracy (*, pred, gold, method='correlation')` (old: `accuracy (hat, mat, distance=False)`)
* `predict_df (*, pred, gold, n=1, method='correlation')` (old: `predict_df (hat, mat, max_guess=1, distance=False, method='cosine')`)
* Adds docstrings to `predict_df`.


**Full Changelog**: https://github.com/quantling/discriminative_lexicon_model/compare/v1.4.3...v2.0.0

- Enables "dlm.performance" after "import discriminative_lexicon_model as dlm".

- Updates some documentations with the LDL class.
- Adds discriminative_lexicon_model.performance.distance_matrix.

This is a python implementation of the Discriminative Lexicon Model and the Linear Discriminative Learning with some helper functions.

This is the first public release, which is published to Pypi as well.

**Full Changelog**: https://github.com/quantling/discriminative_lexicon_model/commits/v1.4.0