This PR seeks to resolve #148.

problem statement

Starting from semTools 0.5-7, monteCarloCI() generates the normal distribution sample with mnormt::rmnorm rather than MASS::mvrnorm. However, there is a slight difference in the values produced by functions from both packages.

• mnormt::rmnorm returns matrix with column names from colname(ACM) (if there is, otherwise X1-XN )
• MASS::nvrnorm returns matrix with column names from names(coefs) forcefully

To avoid this, users have to rewrite their code by assigning column names from names(coefs) manually. I added one line of code to restore the column names from coefs so that users do not need to change their code after updating semTools.

internal behavior

mnormt::rmnorm

> new_unamed <- data.frame(mnormt::rmnorm(n = nRep, mean = coefs, varcov = unnamed))
> names(new_unamed)
[1] "X1" "X2" "X3"
> new_named <- data.frame(mnormt::rmnorm(n = nRep, mean = coefs, varcov = named))
> names(new_named)
[1] "a" "b" "c"

MASS::mvrnorm

> old_unnamed <- data.frame(MASS::mvrnorm(n = nRep, mu = coefs, Sigma = unnamed))
> names(old_unnamed)
[1] "a" "b" "c"
> old_named <- data.frame(MASS::mvrnorm(n = nRep, mu = coefs, Sigma = named))
> names(old_named)
[1] "a" "b" "c"