May Mystery Map Wrap-up

After a series of mystery maps that were cracked within a day, it looks like we finally stumped you! Believe it or not, nobody solved our most recent Mystery Map.  We had some good guesses: several people guessed it was simply distance from Northampton in miles.  This is on the right track, as it offers a plausible explanation for the east/west and north/south gradient originating in Northampton. I was hoping that the two hints that we released would tip someone off: first, we suggested that you check the map more than once; and second, we added dollar signs to the numbers on the map.  Put together the gradient from Northampton, the dollar signs, and a map that changes over time, and you get: a real-time map of Uber fare from Northampton to the 51 largest metro areas* in the US.  That’s right, you can take Uber from Smith College to Portland, Oregon if you so desire. But, it will cost you a pretty penny.  Especially during surge pricing, a dynamic pricing model which increases Uber fares in periods of low supply/high demand which explains why the map changes over time.

Like our other mystery maps, we made this one in CartoDB.  But the real-time aspect makes this one a little more interactive than previous maps.  To add this feature to the map, we took advantage of the Uber API, or Application Program Interface.  An API allows developers (and even wannabe developers such as myself) to talk to applications like Uber with their own applications.  So in this case, my web map makes a request to the Uber API, feeding in the latitude and longitude of Smith’s campus, and Uber sends back the current surge multiplier.  It uses this surge multiplier to calculate a real-time Uber fare to each of the points on the map.  While I find this map titillating, this represents just the tip of the iceberg when it comes to the possibilities that the Uber API presents.  You can see out which services are available in your region (UberX, UberXL, UberBlack, etc), find the locations of drivers near you, and even request a ride without having to open the Uber app itself.

Many (if not most) applications these days have APIs – from commonly used apps like Twitter and Airbnb to programs you may have never heard of, like Building OS, which gathers energy usage data from Smith’s buildings.  All this is to say that there’s so many datasets lurking on the web, waiting to be mapped, analyzed, and mixed with other data.  So while my use of this API is basically just a neat trick, you can probably imagine that APIs could have some real consequence in the realm of big data.  For example, using the Uber API and the Google Maps API, you could examine the impact of Uber on traffic patterns in New York. You could analyze data from the Twitter API with crime data to predict future crimes. Or, you could use the Instagram API with demographic information to assess what people are eating in food deserts.  We helped students use APIs in the Web GIS class that Jon co-taught at UMass this semester, and we can only imagine that APIs will be more commonplace in the SAL as their importance for big data and for GIS grows.

This will be our last Mystery Map until September.  Thanks to those of you who played along, and see you in the fall for some more odysseys in spatial reasoning!



*why 51, you ask? Well, both Jon and I are originally from Rochester, NY, the 51st largest metro area, and hometown pride got the better of us!