Home · Maps · About
Home > BusChat
 

[ Read Responses | Post a New Response | Return to the Index ]
[ First in Thread | Next in Thread ]

 

view flat

Re: Pedestrians aren't the only ones who matter.

Posted by Stephen Bauman on Tue Jan 10 17:15:25 2017, in response to Re: Pedestrians aren't the only ones who matter., posted by JerBear on Tue Jan 10 15:15:29 2017.

edf40wrjww2msgDetailB:detailStr
Stephen, do you have the ability to use Bus Time data from before the Jamaica Ave stop to after the Metropolitan Ave stop?

Anyone has the ability to use the Bus Time data. It's publicly available without filing a FOIL request. There are problems using the Bus Time data.

First, readings are made every 30 seconds. This means that calculated time intervals might be as much as 1 minute off due to sampling errors. That's quite a bit for trying to detect a couple of minutes difference in a 6 to 10 minute duration.

Second, buses frequently are in shadows at the 30 second mark. This means data points are missing. So that 30 second sampling uncertainty can become a minute.

Third, the recorded data is essentially time, latitude and longitude and trip number. It's up to the programmer to figure out the position relative to the bus route. There's a derived output field that gives the "next bus stop." Sometimes it's wrong. The analyzer must take this into account.

Fourth, the driver is supposed to turn the transmitter off when the bus is not on the route. Sometimes they don't which leads to a lot of extra data. Other times they forget to turn the thing back on, when they start the route. This means not all buses along the route are recorded.

Finally, there is a ton of data. I downloaded and placed the data into a database for analysis. It came to over 50 Gig. That was for just for the historical data that's available on the MTA site. I just don't have the disk space to download all the data for a proper analysis within its sampling constraints.

Responses

 Thread is locked Responses disabled

[ Return to the Message Index ]