PlotMplus

R package to plot loading, correlation matrix, scree plot, etc. from MPLUS output.

Functions added so far

• plot_esem
• plot_cor
• plot_loading
• plot_scree

plot_cor

MPLUS report contains the correlation matrix in a lower triangular format. The plot_cor converts it to a full square matrix and use the corrplot package to output a plot of correlation matrix. It uses only the basic features of corrplot but advanced features will be added.

plot_loading

Plots the loading matrix. The loadings are extracted from MPLUS output from the parameters section. Any row with 'BY' is considered as loading. This function returns a ggplot and so can be customize for higher resoluation and other modifications. An example plot: (in low resolution)

plot_esem

In addition to plot the loading the plot_esem function also include the $\beta$ of the $Y=\beta_0 + \beta X + \epsilon$ where $Y$ are the dependent variables and $X$ are the covariates. This is the situation if you want to control for some covariates before fitting the EFA model. In the given sample example, which can be accessed using mplus_out = system.file("extdata", "sample_esem.out", package = "PlotMplus"), we controlled the dependent variables for AGE, SEX, and BMI. Here is the example output from

mplus_out = system.file("extdata", "sample_esem.out", package = "PlotMplus")
plot_esem(model_output = mplus_out, covariates = c('AGE','SEX','BMI'),indep.var = 'CREACT',sort = T)

The last red shaded row represent the parameter estimate of the dependent variable on the factors. In the sample example, the dependent variable was the C-reactive protein.

plot_scree

Use the correlation matrix from MPLUS output and make a simple geom_point ggplot. Some more fancy features will be added.

plot_factor_cor

Use the WITH rows of the parameter estimation output e.g. F2.WITH F1, F2.WITH F2 etc to get correlation matrix and use corrplot to plot it