Clayrs

Added

- Implemented **multiprocessing** via [distex library](https://github.com/erdewit/distex) for the recsys phase!
- `num_cpus` parameter added to each eligible method (`fit()`, `fit_rank()`, `rank()`, `predict()`, etc.)
- Implemented **MAP**, **MAPK** metrics for the _evaluation module_
- Implemented **Bootstrap partitioning** as an additional partitioning technique

Changed
- All algorithms are now able to handle duplicate interactions in the dataset
- In case _data augmentation_ has been performed, _bootstrap partitioning_ has been used, etc.

Changed

- Changed behaviour of all metrics which need to compute popularity
- Now there's no need to re-instantiate them in case multiple runs must be executed

First release of ClayRS!