Raw data and wrapped class objects for the
TransformedOutcome method are stored as class attributes. The wrapped class objects are described in the Usage: modeling section.
Everything else, from processed data to the transformation functions can be accessed as listed below:
up.randomized_search_params # Parameters that are used in `up.randomized_search()`up.grid_search_params # Parameters that are used in `up.grid_search()`up.transform # Outcome transform function.up.untransform # Reverse of outcome transform function.# Data (`y` in any of these can be replaced with `tc` for treatment or `x`).up.transformed_y_train_pred # The predicted uplift.up.transformed_y_train # The transformed outcome.up.y_trainup.y_testup.y # All the `y` data.up.dfup.df_trainup.df_test# Once a model has been created...up.modelup.model_finalup.Q_cgains # 'aqini' or 'qini' can be used in place of 'cgains'up.q1_cgainsup.q2_cgains
The raw data for all evaluation curves can be accessed within any
UpliftEval object (
upev.PLOTTYPE_x # percentileupev.PLOTTYPE_y
where the phrase
PLOTTYPE can be replaced with any of the following:
UpliftEval class objects, they can also be similarly accessed as shown above.
The theoretical maximum curves can also be extracted:
# Overfitting theoretical maximal qini curve.upev.qini_max_x # percentileupev.qini_max_y# "Practical" max curve.upev.qini_pmax_xupev.qini_pmax_y# No sleeping dogs curve.upev.qini_nosdmax_xupev.qini_nosdmax_y
up.train_results_ can be used to plot the qini performance on the training data, as follows: