Had (/am having) some flight adventures today as I slowly make my way toward UC-Davis for the Political Methodology Conference. It started when I slept through my New Jersey Transit airport stop. I've been sleeping on buses, subways, and trains ever since I was a kid, and I believe the last time I failed to wake up for my stop was in elementary school. (In that case, I woke up as the bus doors were closing on my corner, and I alerted the bus driver halfway up the next street.) Back in 2000, I woke up suddenly thinking that the Metro was at my stop, hurriedly jumped out, only to realize that it was the stop before mine. But today my streak ended as I fell victim to a confluence of unfortunate events: I had to hurry to catch the train so I was tired; I didn't intend to fall asleep so I didn't prepare my body to wake up when the train stopped; they weren't announcing stops over the PA; the train was running early so when I woke up and looked at my watch I thought my stop was next, when in fact we were at the stop after mine; the stop after mine was Newark Penn Station, and I failed to realize that Newark Airport comes before Newark Penn Station. Thus, when I finally realized my predicament, the plan was a relatively simple one: get off at Secaucus Junction and take the short cab ride. That plan was foiled when the train zoomed past Secaucus--I had somehow picked one of two off-peak trains that don't stop at Secaucus. Instead, I found myself in Manhattan. Thankfully, since the train was still ahead of schedule, I was able to hop on the first train leaving NYC in the opposite direction, and I arrived at the airport in time to get on the stand-by list, and then finally grab a seat. (My professor, it turned out, was on the same flight--in first class of course!)
My adventures didn't stop there, however. Though we arrived in Minneapolis with plenty of time to make our connection, the flight to Sacramento was cancelled due to mechanical problems. So we scurried over to a counter as asked if there was anyway to get to either Sac., San Fran, or Oakland. The answer was initially "no" (weather problems in Minn. exacerbated the issue), but the friendly woman found a flight to SF through Vegas. San Fran is a good two hours away from Davis, so I enlisted the help of Chicklet as we tried to find suitable ground transport.
Several calls to van services resulted in nothing. And the rental call companies wouldn't let us pick up the car in SF and leave in elsewhere. So we went back to a Northwest counter to turn in our tickets and resign ourselves for an overnight stay in MN. But, the woman checked again, and *now* there were seats available on a later flight to Sacramento! What luck (kind of). So now I'm killing three hours in the twin cities (or, the airport thereof)...if we had stayed overnight I would have tried to get in contact with DogNewTricks, but there's really no point now.
And thus, hopefully (!), ends our saga.
Did you happen to catch Brandt & Freeman's Bayesian time series talk on Saturday morning?
Why, yes, yes I did. Their take home points were to (1) use informative priors, and (2) include error bars on graphs. They were quite adamant about #2 :)