A note on a perfect simulation algorithm for marked Hawkes processes