Testing: add test cases for the different types of objects we want to grade

[nischalshrestha]

We want some coverage of some core objects to test both the output and code checking.

TODOs:

• Python objects (dicts, custom classes like pandas.DataFrame)
• complex expressions that mixes various sub-expressions, for e.g. from this tutorial:

(
  pd.read_csv('data/train.csv')
    .pipe(replace_age_na, pclass_age_map)
    .query('Embarked == "S"')
)

[nischalshrestha]

I think output checking could be modeled after tblcheck where we could provided a summary of issues against target DataFrame