new_boot_ML

Wrapper for creating boot_ML class object.

Produce maximum likelihood estimates of common accuracy statistics
for multiple measurement methods when a gold standard is not available. An R
implementation of the expectation maximization algorithms described in Zhou et
al. (2011) <doi:10.1002/9780470906514> with additional functions for creating
simulated data and visualizing results. Supports binary, ordinal, and
continuous measurement methods.

Corie Drake

emery

Accuracy Statistic Estimation for Imperfect Gold Standards

new_boot_ML function

<dl><dt>v_0</dt>
<dd>MultiMethodMLEstimate S4 object</dd>
<dt>v_star</dt>
<dd>results slot of bootstrapped MultiMethodMLEstimate objects</dd>
<dt>data</dt>
<dd>An <code>n_obs</code> by <code>n_method</code> <code><a href="/link/matrix?package=emery&version=0.7.0" data-mini-rdoc="emery::matrix">matrix</a></code> containing the
observed values for each method. If the dimensions are named, row names will
be used to name each observation (<code>obs_names</code>) and column names will be used
to name each measurement method (<code>method_names</code>).</dd>
<dt>freqs</dt>
<dd>A vector of elements <code>n_obs</code> long representing the number of times
each row in <code>data</code> was observed. Used when <code>data</code> is a summary of unique response
combinations.</dd>
<dt>n_boot</dt>
<dd>number of bootstrap estimates to compute</dd>
<dt>n_study</dt>
<dd>sample size to select for each bootstrap estimate</dd>
<dt>max_iter</dt>
<dd>The maximum number of EM algorithm iterations to compute before
reporting a result.</dd>
<dt>tol</dt>
<dd>The minimum change in statistic estimates needed to continue
iterating the EM algorithm.</dd>
<dt>n_obs</dt>
<dd>Number of observations in data</dd>
<dt>seed</dt>
<dd>optional seed for RNG</dd></dl>

Arguments

Create new boot_ML class object — new_boot_ML

<dl>

<dt>v_0</dt>
<dd>MultiMethodMLEstimate S4 object</dd>


<dt>v_star</dt>
<dd>results slot of bootstrapped MultiMethodMLEstimate objects</dd>


<dt>data</dt>
<dd>An <code>n_obs</code> by <code>n_method</code> <code><a href='https://rdrr.io/r/base/matrix.html'>matrix</a></code> containing the
observed values for each method. If the dimensions are named, row names will
be used to name each observation (<code>obs_names</code>) and column names will be used
to name each measurement method (<code>method_names</code>).</dd>


<dt>freqs</dt>
<dd>A vector of elements <code>n_obs</code> long representing the number of times
each row in <code>data</code> was observed. Used when <code>data</code> is a summary of unique response
combinations.</dd>


<dt>n_boot</dt>
<dd>number of bootstrap estimates to compute</dd>


<dt>n_study</dt>
<dd>sample size to select for each bootstrap estimate</dd>


<dt>max_iter</dt>
<dd>The maximum number of EM algorithm iterations to compute before
reporting a result.</dd>


<dt>tol</dt>
<dd>The minimum change in statistic estimates needed to continue
iterating the EM algorithm.</dd>


<dt>n_obs</dt>
<dd>Number of observations in data</dd>


<dt>seed</dt>
<dd>optional seed for RNG</dd>

</dl>

new_boot_ML: Create new boot_ML class object

Description

Usage

Value

Arguments