Feature/add patsy formulas

[jburos]

Some code refactoring in order to support using patsy formulas for LHS as well as RHS spec. Closes #52

Models can now use the following syntax:

survivalstan.fit_stan_survival_model(
    formula = 'surv(event_status=end_failure, time=t) ~ age + sex',
    data = d,
    model = survivalstan.models.exp_survival_model,
     ...
)

Optional parameters to surv include group= and subject=.

The previous syntax is still also supported:

survivalstan.fit_stan_survival_model(
    event_col = 'end_failure', 
    time_col = 't',
    formula = ' ~ age + sex',
    ...
) 

Moved a fair amount of data-processing into the survivalstan.formulas module, which extends patsy syntax to support the above. For example, the patsy.ModelDesc module includes the stan-data dict which gets passed to pystan.stan for sampling.

This ipynb tests early versions of this code and may help demonstrate how the elements in formulas.py fit together.

[coveralls]

Coverage decreased (-23.7%) to 58.779% when pulling b69d1d5 on feature/add-patsy-formulas into 4bb29b7 on master.

[coveralls]

Coverage decreased (-36.5%) to 45.978% when pulling 3193432 on feature/add-patsy-formulas into 4bb29b7 on master.

[coveralls]

Coverage increased (+1.2%) to 83.673% when pulling 3193432 on feature/add-patsy-formulas into 4bb29b7 on master.

[coveralls]

Coverage increased (+1.6%) to 84.082% when pulling 1946c51 on feature/add-patsy-formulas into 4bb29b7 on master.

[armish]

a total stan/patsy n00b here, so don't feel like I can CR at a higher level and talk about the functionality, but here are a few suggestions/comments regarding the coding style.

[armish]

maybe pull this and all of the other functions above into the SurvivalStanData since I believe they are ony being used in that context. If you think that would make things crowded for a single class, maybe it is worth implementing a utility class that you init with a formula once then can use for further sanity checks, e.g.:

formula = FormulaChecks(formula)
if formula.has_lhs():
   ...

[coveralls]

Changes Unknown when pulling 2dee964 on feature/add-patsy-formulas into ** on dev**.

[julia326]

Some initial comments - let's catch up briefly offline so I can get some background?

[julia326]

add comments explaining what _type and _grouped are?

[TariqAHassan]

@jburos

Any news on this? It would be very cool!

[jburos]

Have migrated this work to PR #77