mvqq: NOTE: this function is replaced and combined into the testdistr function.
Description
This is a simple plotting function designed to help examine
multivariate normality using the (squared) Mahalanobis distance.
Usage
mvqq(dat, use = c("fiml", "pairwise.complete.obs", "complete.obs"),
plot = TRUE)
Arguments
dat
A data frame or matrix of multivariate data to be plotted
use
A character vector indicating how the moments
(means and covariance matrix) should be estimated in the presence of
missing data. The default is to use full information maximum likelihood
based on functions in lavaan.
plot
A logical argument whether to plot the results. Defaults to TRUE.
Value
An invisible list of the density plot, QQ plot, and the data containing
quantiles from the chi-squared distribution. Can be useful to find and remove
multivariate outliers.