transmod {psychonetrics}

Transform between model types

Description

This function allows to transform a model variance–covariance structure from one type to another. Its main uses are to (1) use a Cholesky decomposition to estimate a saturated covariance matrix or GGM, and (2) to transform between conditional (ggm) and marginal associations (cov).

Usage

transmod(x, ..., verbose, keep_computed = FALSE, log = TRUE,
         identify = TRUE)

Arguments

x

A psychonetrics model

...

Named arguments with the new types to use (e.g., between = "ggm" or y = "cov")

verbose

Logical, should messages be printed?

keep_computed

Logical, should the model be stated to be uncomputed adter the transformation? In general, a model does not need to be re-computed as transformed parameters should be at the maximum likelihood estimate.

log

Logical, should a logbook entry be made?

identify

Logical, should the model be identified after transforming?

Details

Transformations are only possible if the model is diagonal (e.g., no partial correlations) or saturated (e.g., all covariances included).

Author(s)

Sacha Epskamp

Examples

# Load bfi data from psych package:
library("psychTools")
data(bfi)

# Also load dplyr for the pipe operator:
library("dplyr")

# Let's take the agreeableness items, and gender:
ConsData <- bfi %>% 
  select(A1:A5, gender) %>% 
  na.omit # Let's remove missingness (otherwise use Estimator = "FIML)

# Define variables:
vars <- names(ConsData)[1:5]

# Model with Cholesky decompositon:
mod <- varcov(ConsData, vars = vars, type = "chol")

# Run model:
mod <- mod %>% runmodel

# Transform to GGM:
mod_trans <- transmod(mod, type = "ggm") %>% runmodel
# Note: runmodel often not needed

# Obtain thresholded GGM:
getmatrix(mod_trans, "omega", threshold = TRUE)