Starting a thread for noteworthy changes or issues with data including the GTFS feeds.
Not sure how long this has been the case but tonight’s my first time seeing rail replacement on GTFS Realtime. Big win for minimising disruption.
5 Likes
I guess this is a good thread to talk about Fluvial 
(I’ve been kinda shadowy on it for a while to avoid pseud/realname crossover but that’s no longer an issue.)
For the folks who haven’t seen it, it is a visualiser for origin-destination patronage data along a route, like this:
Usually, GTFS data is used to establish a stop order. This is simple until it isn’t — we have all sorts of route variations, tails and part-time expresses, some of our routes do loops, some of them do figure eights…
SEQ’s patronage data comes aggregated by the month, but it is possible to filter it a little by time of day (weekday AM peak, weekday middle, weekday PM peak, weeknights, weekends). It’s also broken down by route and origin/destination.
However, QR is all bundled up as a single “Rail” route because the readers are on the platform rather than on the vehicle. I’ve been trying to work through the best way to deal with that (from configuration and auto-discovery) for… three years now.
1 Like
How hard would it be to use this data to model a redesigned BRT style network. Yes of course changing routes will change behaviour but it would be an interesting analysis.
I wouldn’t rely on it too heavily for major network changes.
You don’t measure cross-river bridge demand by the number of people swimming!
If a route is well patronised despite being low frequency and it’s where you’d put a BRT route anyway, then sure, that’s evidence for an upgrade.
You can also use this data to think about patronage along a corridor.
E.G. for the SE Busway some people catch the 130 to Queen St but some people catch a P-rocket to the city; if the P-rocket didn’t exist, all that demand would have to go on the 130.
Yer it was that consolidation # that I was interested in. To see the patronage levels through just consolidating/de-duplicating.
Once the 124/125 changes happen that’ll be good data I reckon
1 Like