shared_args

Arguments that are shared across functions

internal

Causal Distillation Tree (CDT) is a novel machine learning method
for estimating interpretable subgroups with heterogeneous treatment effects.
CDT allows researchers to fit any machine learning model (or metalearner) to
estimate heterogeneous treatment effects for each individual, and then
"distills" these predicted heterogeneous treatment effects into
interpretable subgroups by fitting an ordinary decision tree to predict the
previously-estimated heterogeneous treatment effects. This package
provides tools to estimate causal distillation trees (CDT), as detailed in
Huang, Tang, and Kenney (2025) <doi:10.48550/arXiv.2502.07275>.

Tiffany Tang

causalDT

Causal Distillation Trees

Melody Huang

Ana Kenney

shared_args function

<dl><dt>X</dt>
<dd>A tibble, data.frame, or matrix of covariates.</dd>
<dt>Y</dt>
<dd>A vector of outcomes.</dd>
<dt>y</dt>
<dd>A vector of responses to predict.</dd>
<dt>Z</dt>
<dd>A vector of treatments.</dd>
<dt>W</dt>
<dd>A vector of weights corresponding to treatment propensities.</dd>
<dt>rpart_control</dt>
<dd>A list of control parameters for the <code>rpart</code> algorithm.
See <code>? rpart.control</code> for details.</dd>
<dt>rpart_fit</dt>
<dd>An <code>rpart</code> object.</dd>
<dt>party_fit</dt>
<dd>A <code>party</code> object.</dd>
<dt>prune</dt>
<dd>Method for pruning the tree. Default is <code>"none"</code>. Options
are <code>"none"</code>, <code>"min"</code>, and <code>"1se"</code>. If <code>"min"</code>, the
tree is pruned using the complexity threshold which minimizes the
cross-validation error. If <code>"1se"</code>, the tree is pruned using the
largest complexity threshold which yields a cross-vaidation error within
one standard error of the minimum. If <code>"none"</code>, the tree is not
pruned.</dd>
<dt>rpart_prune</dt>
<dd>Method for pruning the tree. Default is <code>"none"</code>.
Options are <code>"none"</code>, <code>"min"</code>, and <code>"1se"</code>. If <code>"min"</code>,
the tree is pruned using the complexity threshold which minimizes the
cross-validation error. If <code>"1se"</code>, the tree is pruned using the
largest complexity threshold which yields a cross-vaidation error within
one standard error of the minimum. If <code>"none"</code>, the tree is not
pruned.</dd></dl>

Arguments

Arguments that are shared across functions — shared_args

<dl>

<dt>X</dt>
<dd>A tibble, data.frame, or matrix of covariates.</dd>


<dt>Y</dt>
<dd>A vector of outcomes.</dd>


<dt>y</dt>
<dd>A vector of responses to predict.</dd>


<dt>Z</dt>
<dd>A vector of treatments.</dd>


<dt>W</dt>
<dd>A vector of weights corresponding to treatment propensities.</dd>


<dt>rpart_control</dt>
<dd>A list of control parameters for the <code>rpart</code> algorithm.
See <code>? rpart.control</code> for details.</dd>


<dt>rpart_fit</dt>
<dd>An <code>rpart</code> object.</dd>


<dt>party_fit</dt>
<dd>A <code>party</code> object.</dd>


<dt>prune</dt>
<dd>Method for pruning the tree. Default is <code>"none"</code>. Options
are <code>"none"</code>, <code>"min"</code>, and <code>"1se"</code>. If <code>"min"</code>, the
tree is pruned using the complexity threshold which minimizes the
cross-validation error. If <code>"1se"</code>, the tree is pruned using the
largest complexity threshold which yields a cross-vaidation error within
one standard error of the minimum. If <code>"none"</code>, the tree is not
pruned.</dd>


<dt>rpart_prune</dt>
<dd>Method for pruning the tree. Default is <code>"none"</code>.
Options are <code>"none"</code>, <code>"min"</code>, and <code>"1se"</code>. If <code>"min"</code>,
the tree is pruned using the complexity threshold which minimizes the
cross-validation error. If <code>"1se"</code>, the tree is pruned using the
largest complexity threshold which yields a cross-vaidation error within
one standard error of the minimum. If <code>"none"</code>, the tree is not
pruned.</dd>

</dl>

shared_args: Arguments that are shared across functions

Description

Arguments