Add to the pipeline the creation of the files mec_vectorized_1 and mec_vectorized_2

[dancoster]

Add an option to use only the top K (=200,300) medications that are common in the dataset per patient (not total number of adminstration).

[PavanReddy28]

Implemented as part of pull request - https://github.com/dancoster/DrugLab/pull/16