ndr_core: Core function of 'ndr_learner'. It executes steps 2 and 4 of...

In MCKnaus/causalDML: Causal Double Machine Learning

Core function of ndr_learner. It executes steps 2 and 4 of algorithms 1 and 2 in Knaus (2020).

Description

Core function of ndr_learner. It executes steps 2 and 4 of algorithms 1 and 2 in Knaus (2020).

Usage

ndr_core(
  ml,
  delta,
  y,
  x,
  w_mat,
  m_mat,
  e_mat,
  cf_mat,
  xnew = NULL,
  nfolds = 5,
  path = NULL,
  quiet = TRUE
)

Arguments

ml	List of methods to be used in ensemble estimation. Methods can be created by create_method.
delta	vector of doubly robust score. E.g. create via ATE_dml.
y	Vector of variable to be predicted.
x	Matrix of covariates.
w_mat	Logical matrix of treatment indicators (n x T+1). For example created by prep_w_mat.
m_mat	n x T+1 matrix with fitted outcome values.
e_mat	n x T+1 matrix with propensity scores.
cf_mat	Logical matrix with k columns of indicators representing the different folds (for example created by prep_cf_mat).
xnew	Covariate matrix of test sample
nfolds	Number of folds used in cross-validation of ensemble weights (default nfolds=5)
path	Optional path to save the ensemble objects for later processing. Saved as Ensemble_Yi where i is the number of the treatment in multiple treatment settings.
quiet	If FALSE, ensemble estimators print method that is currently running.

Value

n x 2 matrix containing DR- and NDR-learner predictions.

References

• Knaus, M. C. (2020). Double machine learning based program evaluation under unconfoundedness. arXiv preprint arXiv:2003.03191.http://arxiv.org/abs/2003.03191

MCKnaus/causalDML documentation built on Aug. 19, 2023, 5:47 p.m.