get_average_effect.rst

get_average_effect

Estimating an average effect of the test set.

Parameters:	y_test: numpy array   Actual y values. t_test: numpy array   Actual treatment values. y_pred: numpy array   Predicted y values by uplift model. test_share: float   Share of the test data which will be taken for estimating an average effect.
Returns:	average effect: float   Average effect on the test set.

Examples

from pyuplift.metrics import get_average_effect
...
model.fit(X_train, y_train, t_train)
y_pred = model.predict(X_test)
effect = get_average_effect(y_test, t_test, y_pred, test_share)
print(effect)