semPlotModel-class: Class `"semPlotModel"`

Description

Representation of SEM models, can be used by semPaths, semCors and semSyntax. See semPlotModel-edit for utility functions on how to edit this model.

Arguments

Objects from the Class

Objects can be created by calls of the form new("semPlotModel", ...).

Slots

Pars:

Object of class "data.frame" indicating the parameters used in the SEM model. this must contain the following elements, in order:

label

The name of the parameter, used as edge label in the graph.

lhs

Name of the variable on the left hand side of the path.

edge

String as indicator of the edge. This can be one of the following:

->: Factor loading

~>

Regression. The same as '->' in that it results in a directed edge from the left hand side to the right hand side, but '~>' differs in that if the right hand side is manifest and the left hand side is an exogenous latent the right hand side is interpreted as an endogenous variable rather than an exogenous variable.

<->

(co)variance

int

intercept, The left hand side should be "" and the right hand side indicates the variable to which the intercept belongs.

--

Undirected edge. Only used as dummy encoding and in cases the parameter can not be interpreted (usually this indicates something that is not yet supported)

rhs

Name of the variable on the left hand side of the path.

est

Parameter estimate.

est

Standardized parameter estimate.

group

Character of the name of the group the parameter belongs to.

fixed

Logical indicating if the parameter is fixed.

par

Parameter number. 0 indicates the parameter is fixed and parameters with the same parameter number are constrained to be equal.

knot

Knot number. 0 indicates the edge is not knotted and edges with the same knot number are knotted together. Only used to indicate interactions in 'lm' models and can be omitted.

Vars:

Object of class "data.frame" indicating the variables used in the SEM model. Must have the following elements:

name: Name of the variable

manifest

Logical indicating if the variable is manifest

exogenous

Logical indicating if the variable is exogenous. If NA

(the default) semPaths will attempt to detect which variables are exogenous.

Thresholds:

Object of class "data.frame" indicating the thresholds in the SEM model. It is the same as Pars except it does not have the elements 'edge' and 'rhs'.

Computed:

Object of class "logical" indicating if the SEM model was computed or if the object only indicates a structure.

ObsCovs:

Object of class "list" containing observed covariance matrices for each group. If available.

ImpCovs:

Object of class "list" containing implied covariance matrices for each group. If available.

Original:

Object of class "list" containing the original object used as input (or multiple objects if the '+' operator was used to combine objects.)

Methods

No methods defined with class "semPlotModel" in the signature.

Author

Sacha Epskamp <mail@sachaepskamp.com>

Examples

Run this code

showClass("semPlotModel")

Run the code above in your browser using DataLab