additions to sim_cat()

[Aariq]

Originally sim_cat() was a place to start a dataframe. But I could do more to make it generally useful for generating factors or categorical data.

Enhancements to sim_cat():

• can accept dataframe optionally as argument

• can create multiple categorical variables as crossed, nested, or random

[Aariq]

See mclust::cross for an example of crossed data

[Aariq]

Also see purrr::cross_df