splitopt

############################################################################################

internal

It provides an interesting solution for handling a high number
of segmentation variables in partial least squares structural equation
modeling. The package implements the "Pathmox" algorithm (Lamberti, Sanchez,
and Aluja,(2016)<doi:10.1002/asmb.2168>) including the F-coefficient
test (Lamberti, Sanchez, and Aluja,(2017)<doi:10.1002/asmb.2270>)
to detect the path coefficients responsible for the identified differences).
The package also allows running the hybrid multi-group approach (Lamberti (2021)
<doi:10.1007/s11135-021-01096-9>).

Giuseppe Lamberti

genpathmox

Pathmox Approach Segmentation Tree Analysis

splitopt function

<dl><dt>x</dt>
<dd>matrix or data frame containing the manifest variables.</dd>
<dt>inner</dt>
<dd>a square (lower triangular) boolean matrix representing
the inner model (i.e. the path relationships between latent variables)</dd>
<dt>.model</dt>
<dd>A description of the user-specified model.</dd>
<dt>.scheme</dt>
<dd>string indicating the type of inner weighting
scheme. Possible values are <code>"centroid"</code>, <code>"factorial"</code>, or
<code>"path"</code>.</dd>
<dt>.consistent</dt>
<dd>Logical. Should composite/proxy correlations be disattenuated
to yield consistent loadings and path estimates if at least one of the construct
is modeled as a common factor. Defaults to TRUE.</dd>
<dt>splits</dt>
<dd>vector indicating the binary partition.</dd>
<dt>fact</dt>
<dd>vector indicating the categorical variable.</dd>
<dt>size_candidate</dt>
<dd>number indicating the minimum threshold for a node.</dd></dl>

Arguments

Best partition for a specific segmentation variable — splitopt

<dl>

<dt>x</dt>
<dd>matrix or data frame containing the manifest variables.</dd>


<dt>inner</dt>
<dd>a square (lower triangular) boolean matrix representing
the inner model (i.e. the path relationships between latent variables)</dd>


<dt>.model</dt>
<dd>A description of the user-specified model.</dd>


<dt>.scheme</dt>
<dd>string indicating the type of inner weighting
scheme. Possible values are <code>"centroid"</code>, <code>"factorial"</code>, or
<code>"path"</code>.</dd>


<dt>.consistent</dt>
<dd>Logical. Should composite/proxy correlations be disattenuated
to yield consistent loadings and path estimates if at least one of the construct
is modeled as a common factor. Defaults to TRUE.</dd>


<dt>splits</dt>
<dd>vector indicating the binary partition.</dd>


<dt>fact</dt>
<dd>vector indicating the categorical variable.</dd>


<dt>size_candidate</dt>
<dd>number indicating the minimum threshold for a node.</dd>

</dl>

splitopt: Best partition for a specific segmentation variable

Description

Usage

Value

Arguments

Details