Do you have an Uber Problem ? I do. I’d catch an Uber to the toilet if I could.
Recently though, they released anonymised dumps of their travel datasets so I thought I’d dig into them to get a sense of how people are using the service in some towns I know and love:
- My home town of Brisbane;
- My other hometown of Sydney; and
- London – a place where I lived, worked, and continue to visit.
Below are the initial findings done mostly with bash scripts creating javascript calling out to Google Maps APIs. The intention being to arrive at a quick view of the size of shape of the data before diving into the next part which would be putting it into Tableau and looking at some graph or network representations, or maybe just doing some simpler colour-coding, weighting and interactivity.
Brisbane
Although it’s thriving now and set to host the Olympics in 2032, back in day the phrase “a jumped-up country town” was often used. Partially this was true because I still remember walking the city streets following after-school sports training and trying to get home on the final run of one of the three bus routes available. It was probably fair to say that the place was a ghost town.
So what does the Uber data tells about the city and it’s denizens today ? Uber released the data in fairly aggregated form and sanitized for personally identifiable information. So I thought I’d choose a recent, busy period to give us an upper bounds – November 2018 to January 2019:
Sydney
Same time-frame, much larger and older city than Brisvegas:
London
As mentioned I’ve lived and worked here and had a ball. Check out the graph though – strikingly radial. I guess these are the townships the tube forget ? Or where people cash-in together to the avoid the possible hassles of same ? Very eerie to see such regularity amongst such a large, mobile population. Hopefully some Tableau analysis to follow but here’s the first look:
That’s it for the intial look – hopefully more to follow later.