QmatrixToBlatentSyntax {blatent}

Convert a rectangular Q-matrix into blatent model syntax

Description

Converts a rectangular Q-matrix into blatent model syntax. Q-matrix must have observed variables listed across columns and latent variables listed across rows.

Usage

QmatrixToBlatentSyntax(
  Qmatrix,
  observedVariables = "rownames",
  latentVariables = "colnames",
  lvDist = "joint"
)

Arguments

Qmatrix

A data frame or matrix containing a Q-matrix.

observedVariables

If Qmatrix is data.frame, the variable in the data frame that has the names of the observed variables. Defaults to "rownames", which uses rownames(Qmatrix) and works for data.frame or matrix types of Qmatrix.

latentVariables

A vector of the variable or column names of the latent variables. Defaults to "colnames", which uses colnames(Qmatrix) and works for data.frame or matrix types of Qmatrix.

lvDist

A character that indicates the type of latent variable distribution to be used. "joint" for joint distributions (multivariate Bernoulli) or "univariate" for univariate Bernoulli using BayesNets parameterization. The latter also builds blatent syntax for the BayesNets model terms.

Value

A character vector containing blatent model syntax.

Examples

# Example 1: Joint distribution using data.frame
# empty data.frame

exampleQmatrixDF = data.frame(matrix(data = 0, nrow = 10, ncol = 3))

# name columns of Qmatrix

names(exampleQmatrixDF) = c("observedVariableName", "Attribute1", "Attribute2")

# names of observed variables

exampleQmatrixDF[1:10, "observedVariableName"] = paste0("Item",1:10)

# Entries for Qmatrix

exampleQmatrixDF[1:5,"Attribute1"] = 1
exampleQmatrixDF[3:10,"Attribute2"] = 1

# produce blatentSyntax using QmatrixToBlatentSyntax() function

blatentSyntaxJoint = QmatrixToBlatentSyntax(
   Qmatrix = exampleQmatrixDF,
   observedVariables = "observedVariableName",
   latentVariables = c("Attribute1", "Attribute2"),
   lvDist = "joint"
 )
cat(blatentSyntaxJoint)

# Example 2: Univariate distributions using matrix
# empty data.frame

exampleQmatrixM = matrix(data = 0, nrow = 10, ncol = 2)

# name columns of Qmatrix as latent variable names

colnames(exampleQmatrixM) = c("Attribute1", "Attribute2")

# name rows of Qmatrix as observed variable names

rownames(exampleQmatrixM) = paste0("Item",1:10)

# Entries for Qmatrix

exampleQmatrixM[1:5,"Attribute1"] = 1
exampleQmatrixM[3:10,"Attribute2"] = 1

# produce blatentSyntax using QmatrixToBlatentSyntax() function
#  (with default options for observedVariables and latentVariables)

blatentSyntaxM = QmatrixToBlatentSyntax(Qmatrix = exampleQmatrixM,  lvDist = "univariate")
cat(blatentSyntaxM)