cleverly.interventions.LinearQuantileRegression

class cleverly.interventions.LinearQuantileRegression(features: tuple[str, ...])

A declared linear quantile regression: the boundary beta' (1, features).

The finite-dimensional boundary learner of SupportBoundary. It fits the level-q check loss of the dose on an intercept and the named columns of the policy frame, on the units at risk. \(\beta\) solves an estimating equation, so the population target has an influence curve, whether or not the regression is correctly specified (the pseudo-true \(\beta\)).

Parameters:

featurestuple of str

Columns of the policy frame. An intercept is added.

Methods

Attributes

LinearQuantileRegression.features: tuple[str, ...]