Skip to contents

A small built-in learner for dml_backend = "custom" examples and tests. It fits one least-squares model per (d, post) cell and predicts all four outcome-regression nuisance functions for each test observation.

Usage

compdid_cell_or(dp_train, dp_test, ...)

Arguments

dp_train

Training dp object.

dp_test

Test dp object.

...

Unused.

Value

An n_test x 4 matrix of outcome-regression predictions.