Create simulated stream of participants

[JamesPHoughton]

A stream of participants arrives at the door. Each participant has attributes,

[
  {
    party: "democrat",
    age: 34,
    gender: f,
    arrivalTime: 14,
    dropoutTime: 25
 },
 {
    party: "democrat",
    age: 19,
    gender: m,
    arrivalTime: 18,
    dropoutTime: 30
 },
...
]

The task is to generate arbitrary streams like this with reasonable distributional properties:

• [ ] party should be 35% R, 45% D 20% I
• [ ] age should be truncnorm with min 19, max 80, mean 45, std 15?
• [ ] gender 55%m
• [ ] arrivalTime is lognormal with mean at 20 seconds and a long tail
• [ ] departureTime is always greater than arrival time, probably arrivaltime + exponential w/ 60 second mean
• [ ] sorted by arrival time