pive_weight_maker

This is a generic function taking pseudo-instrument,
pseudo-treatment and the transformation matrix as inputs and returning
outcome weights

Many treatment effect estimators can be written as weighted outcomes.
These weights have established use cases like checking covariate balancing via packages like 'cobalt'.
This package takes the original estimator objects and outputs these outcome weights.
It builds on the general framework of Knaus (2024) <doi:10.48550/arXiv.2411.11559>.
This version is compatible with the 'grf' package and provides an internal implementation of Double Machine Learning.

Knaus Michael C.

OutcomeWeights

Outcome Weights of Treatment Effect Estimators

Michael C. Knaus

Henri Pfleiderer

pive_weight_maker function

<dl><dt>Z.tilde</dt>
<dd>Numeric vector of pseudo-instrument outcomes.</dd>
<dt>D.tilde</dt>
<dd>Numeric vector of pseudo-treatment.</dd>
<dt>T_mat</dt>
<dd>Transformation matrix</dd></dl>

Arguments

Outcome weights maker for pseudo-IV estimators. — pive_weight_maker

<dl>

<dt>Z.tilde</dt>
<dd>Numeric vector of pseudo-instrument outcomes.</dd>


<dt>D.tilde</dt>
<dd>Numeric vector of pseudo-treatment.</dd>


<dt>T_mat</dt>
<dd>Transformation matrix</dd>

</dl>

pive_weight_maker: Outcome weights maker for pseudo-IV estimators.

Description

Usage

Value

Arguments

References