• This entire project was developed in ipython; we recommend you run it in ipython as well • It relies on the following packages: - pandas - numpy/scipy - sklearn - gensim • All of these can be installed via pip in a virtual environment. that is probably your best bet.

• All data is located in ./data, including both the original (raw, as received) json files and our curated 'pandas' dataframes • Relevant information should be contained in json_obj['hits']['hits']['_source'], which will be a list of dicts representing calendar events, etc.

• original: es-calendar.json, curated pandas dataframe in calendar.df

• original: es-activity.json, curated pandas dataframe in activity.df •

• To initialize:

> from Preprocess import Preprocess
> pre = Preprocess ()

• To correctly format an object containing calendar events: [ce_obj is either a pandas dataframe or raw json from their database]

> ce_df = pre.preprocess_ce (ce_obj)

• To correctly format an object containing activities [a_obj is either a pandas dataframe or a raw json from their database]

> a_df = pre.preprocess_a (a_obj)