Sigma_transformation {modgo}R Documentation

Correlation of transformed variables

Description

This function is used internally by modgo. It finishes the computation of the correlation matrix of the transformed variables, which are assumed to follow a multivariate normal distribution. It computes the correlations involving at least one categorical variable. For this purpose the biserial, tetrachoric, polyserial and polychoric correlations are used.

Usage

Sigma_transformation(
  data,
  data_z,
  Sigma,
  variables,
  bin_variables = c(),
  categ_variables = c()
)

Arguments

data

a data frame with original variables.

data_z

data frame with transformed variables.

Sigma

A numeric square matrix.

variables

variables a character vector indicating which columns of data should be used.

bin_variables

a character vector listing the binary variables.

categ_variables

a character vector listing the ordinal categorical variables.

Value

A numeric matrix with correlation values.

Author(s)

Francisco M. Ojeda, George Koliopanos


[Package modgo version 1.0.1 Index]