impute_internal

This is the work horse function that implements most of the functionality of impute.
See the user level function <code>impute()</code> for further details.

Implements standard and reference based multiple imputation methods for continuous
longitudinal endpoints (Gower-Page et al. (2022) <doi:10.21105/joss.04251>). In particular,
this package supports deterministic conditional mean imputation and jackknifing as described
in Wolbers et al. (2022) <doi:10.1002/pst.2234>, Bayesian multiple imputation as described
in Carpenter et al. (2013) <doi:10.1080/10543406.2013.834911>, and bootstrapped maximum
likelihood imputation as described in von Hippel and Bartlett (2021) <doi: 10.1214/20-STS793>.

Isaac Gravestock

rbmi

Reference Based Multiple Imputation

Craig Gower-Page

Alessandro Noci

Marcel Wolbers

Daniel Sabanes Bove

F. Hoffmann-La Roche AG 

impute_internal function

<dl><dt>draws</dt>
<dd>A <code>draws</code> object created by <code>draws()</code>.</dd>
<dt>references</dt>
<dd>A named vector. Identifies the references to be used for reference-based
imputation methods. Should be of the form <code>c("Group1" = "Reference1", "Group2" = "Reference2")</code>.
If <code>NULL</code> (default), the references are assumed to be of the form
<code>c("Group1" = "Group1", "Group2" = "Group2")</code>. This argument cannot be <code>NULL</code> if
an imputation strategy (as defined by <code>data_ice[[vars$strategy]]</code> in the call to draws) other than <code>MAR</code> is set.</dd>
<dt>update_strategy</dt>
<dd>An optional <code>data.frame</code>. Updates the imputation method that was
originally set via the <code>data_ice</code> option in <code>draws()</code>. See the details section for more
information.</dd>
<dt>strategies</dt>
<dd>A named list of functions. Defines the imputation functions to be used.
The names of the list should mirror the values specified in <code>strategy</code> column of <code>data_ice</code>.
Default = <code>getStrategies()</code>. See <code>getStrategies()</code> for more details.</dd>
<dt>condmean</dt>
<dd>logical. If TRUE will impute using the conditional mean values, if values
will impute by taking a random draw from the multivariate normal distribution.</dd></dl>

Arguments

Create imputed datasets — impute_internal

<dl>

<dt>draws</dt>
<dd>A <code>draws</code> object created by <code>draws()</code>.</dd>


<dt>references</dt>
<dd>A named vector. Identifies the references to be used for reference-based
imputation methods. Should be of the form <code>c("Group1" = "Reference1", "Group2" = "Reference2")</code>.
If <code>NULL</code> (default), the references are assumed to be of the form
<code>c("Group1" = "Group1", "Group2" = "Group2")</code>. This argument cannot be <code>NULL</code> if
an imputation strategy (as defined by <code>data_ice[[vars$strategy]]</code> in the call to draws) other than <code>MAR</code> is set.</dd>


<dt>update_strategy</dt>
<dd>An optional <code>data.frame</code>. Updates the imputation method that was
originally set via the <code>data_ice</code> option in <code>draws()</code>. See the details section for more
information.</dd>


<dt>strategies</dt>
<dd>A named list of functions. Defines the imputation functions to be used.
The names of the list should mirror the values specified in <code>strategy</code> column of <code>data_ice</code>.
Default = <code>getStrategies()</code>. See <code>getStrategies()</code> for more details.</dd>


<dt>condmean</dt>
<dd>logical. If TRUE will impute using the conditional mean values, if values
will impute by taking a random draw from the multivariate normal distribution.</dd>

</dl>

impute_internal: Create imputed datasets

Description

Usage

Arguments