Benchmarking mclust and MPlus

In tidyLPA: Easily Carry Out Latent Profile Analysis (LPA) Using Open-Source or Commercial Software

Purpose

The purpose of this vignette is to facilitate benchmarking the output from the use of two packages, MPlus (via the MplusAutomation package) and mclust (via the mclust package). As CRAN does not allow for MPlus to be used, this vignette is not rendered, but can be rendered on your own computer as long as you have MPlus (and MplusAutomation) installed. Please note that MPlus must be installed on your computer for this vignette to run.

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>", 
  eval = FALSE
)

Using mclust

Just one fit

library(tidyLPA)
library(dplyr)

iris %>% 
  select(Sepal.Length, Sepal.Width, Petal.Length, Petal.Width) %>% 
  estimate_profiles(n_profiles = 2) %>% 
  get_fit() %>% 
  as.data.frame()

Using MPlus

Just one fit

iris %>% 
  select(Sepal.Length, Sepal.Width, Petal.Length, Petal.Width) %>% 
  estimate_profiles(n_profiles = 2, package = "mplus") %>% 
  get_fit() %>% 
  as.data.frame()

Using mclust

Many fits

library(tidyLPA)
library(dplyr)

iris %>% 
  select(Sepal.Length, Sepal.Width, Petal.Length, Petal.Width) %>% 
  estimate_profiles(n_profiles = 2:4, models = c(1, 2, 3, 6)) %>% 
  get_fit() %>% 
  as.data.frame()

Using MPlus

Many fits

iris %>% 
  select(Sepal.Length, Sepal.Width, Petal.Length, Petal.Width) %>% 
  estimate_profiles(n_profiles = 2:4, models = c(1, 2, 3, 6),
                    package = "mplus") %>% 
  get_fit() %>% 
  as.data.frame()

tidyLPA documentation built on Nov. 17, 2021, 5:07 p.m.