Skip to content
This repository has been archived by the owner on Jul 29, 2024. It is now read-only.

Latest commit

 

History

History
25 lines (16 loc) · 1.07 KB

quick-start.md

File metadata and controls

25 lines (16 loc) · 1.07 KB

Quick start

Install pylift:

pip install pylift

To start, you simply need a pandas.DataFrame with a treatment column of 0s and 1s (0 for control, 1 for test) and a numeric outcome column. Implementation can be as simple as follows:

from pylift import TransformedOutcome
up = TransformedOutcome(df1, col_treatment='Treatment', col_outcome='Converted')

up.randomized_search()
up.fit(**up.rand_search_.best_params_)

up.plot(plot_type='aqini', show_theoretical_max=True)
print(up.test_results_.Q_aqini)

up.fit() can also be passed a flag productionize=True, which when True will create a productionizable model trained over the entire data set, stored in self.model_final (though it is contentious whether it's safe to use a model that has not been evaluated in production -- if you have enough data, it may be prudent not to). This can then be pickled with self.model_final.to_pickle(PATH), as usually done with sklearn-style models.

A fully-fledged example of pylift in action can be found in pylift/examples/simulated_data/sample.ipynb.