What is the advantage of using the pandas function get$\_$dummies$()$ over the one$-$hot$-$encoder class in sklearn.preprocessing?

The one$-$hot$-$encoder class is easier to use

The pandas function get$\_$dummies$()$ is much faster

The one$-$hot$-$encoder class allows for more customization

The pandas function get$\_$dummies$()$ is much easier to use