compute_marg_PR_nested_reg

The array version of this function (compute_marg_PR_nested_reg_array) is used in plot_etiology_regression

Provides functions to specify, fit and visualize
nested partially-latent class models (
Wu, Deloria-Knoll, Hammitt, and Zeger (2016) <doi:10.1111/rssc.12101>;
Wu, Deloria-Knoll, and Zeger (2017) <doi:10.1093/biostatistics/kxw037>;
Wu and Chen (2021) <doi:10.1002/sim.8804>) for
inference of population disease etiology and individual diagnosis. In the motivating
Pneumonia Etiology Research for Child Health (PERCH) study, because both quantities
of interest sum to one hundred percent, the PERCH scientists frequently refer to
them as population etiology pie and individual etiology pie, hence the name of the package.

Zhenke Wu

baker

"Nested Partially Latent Class Models"

Scott Zeger

John Muschelli

Irena Chen

compute_marg_PR_nested_reg function

<dl><dt>ThetaBS</dt>
<dd>True positive rates for <code>JBrS</code> measures (rows) among <code>K</code> subclasses (columns)</dd>
<dt>PsiBS</dt>
<dd>False positive rates; dimension same as above</dd>
<dt>pEti_mat</dt>
<dd>a matrix of etiology pies for <code>N</code> subjects (rows) and <code>Jcause</code> causes (columns)
rows sum to ones.</dd>
<dt>subwt_mat</dt>
<dd>a matrix of subclass weights for cases and controls. <code>N</code> by <code>K</code>. Rows sum
to ones.</dd>
<dt>case</dt>
<dd>a N-vector of <code>1</code>s (cases) and <code>0</code>s (controls)</dd>
<dt>template</dt>
<dd>a binary matrix with <code>Jcause+1</code> rows (<code>Jcause</code> classes of cases and <code>1</code> class of controls)
and <code>JBrS</code> columns for the Bronze-standard measurement (say, pick one type/slice).
The ones in each row indicate the measurements that will show up more frequently in cases given the cause.</dd></dl>

Arguments

compute positive rates for nested model with subclass mixing weights that are the same across
Jcause classes for each person (people may have different weights.) — compute_marg_PR_nested_reg

<dl>

<dt>ThetaBS</dt>
<dd>True positive rates for <code>JBrS</code> measures (rows) among <code>K</code> subclasses (columns)</dd>


<dt>PsiBS</dt>
<dd>False positive rates; dimension same as above</dd>


<dt>pEti_mat</dt>
<dd>a matrix of etiology pies for <code>N</code> subjects (rows) and <code>Jcause</code> causes (columns)
rows sum to ones.</dd>


<dt>subwt_mat</dt>
<dd>a matrix of subclass weights for cases and controls. <code>N</code> by <code>K</code>. Rows sum
to ones.</dd>


<dt>case</dt>
<dd>a N-vector of <code>1</code>s (cases) and <code>0</code>s (controls)</dd>


<dt>template</dt>
<dd>a binary matrix with <code>Jcause+1</code> rows (<code>Jcause</code> classes of cases and <code>1</code> class of controls)
and <code>JBrS</code> columns for the Bronze-standard measurement (say, pick one type/slice).
The ones in each row indicate the measurements that will show up more frequently in cases given the cause.</dd>

</dl>

compute positive rates for nested model with subclass mixing weights that are the same across
<code>Jcause</code> classes for each person (people may have different weights.)

compute_marg_PR_nested_reg: compute positive rates for nested model with subclass mixing weights that are the same across `Jcause` classes for each person (people may have different weights.)

Description

Usage

Value

Arguments