The same irregularity in size and shape that this map fixes also bites it. Now, even a modest study of the tile grid version will show its issue: uhhh… states are in funny places. Yes, I am deficient in the Plains, Upper Midwest, and Deep South, but the numbers aren’t quite so bad as those big, white states on the choropleth suggest. What’s more, from a data visualization perspective, now each state is the same size and the distribution is a bit fairer. Even Alaska and Hawaii get handled a bit more elegantly. is still there, including the classic markers of New England, Florida, Texas, and even Michigan. You can see that the rough shape of the U.S. Here is the same data on a tile grid map of my own construction: All in all, I have spent considerable time in a large number of states, but, because they are mostly small and packed into the northeast, the number seems smaller than it it is. It so happens I am a Northeasterner, and I have spent much of my time in small states. The map is familiar and we can pick out the data, but it does illustrate the same issue as the London one: the irregular sizes of states distort the view. To illustrate the idea using a map more familiar to many in the audience, consider a simple dataset: the time I have spent in each U.S. The short answer is, it works, but not without some trade-offs. My first thought was, obviously, “cool – but could this work for the U.S.?” I put pencil to paper. Can This Idea Be More Broadly Applicable? It is a set of equal shapes arranged to mimic an actual map and better display data that might otherwise get distorted or limited by a classic choropleth view. The visualization above is a rich tile grid map. Even more, the squares allowed for inclusion of different sorts of data displays, making the map a nice small multiple visualization. It isn’t clear what the data represents, but you get the idea. In this version, every borough was equal, and one could better mentally count numbers in categories if shading represented data. Their solution was to abstract the city into a set of equal squares, save a few in the center that encapsulate the iconic path of the Thames River for reference. In aggregate, the view could distort the viewers’ perceptions of the number of boroughs in different categories. A dark color in a small borough could get lost amid lighter colors for a bigger borough. Because of the differences in size and odd distribution, shading could tell a muddled story. Many are small and all are irregular, leaving little space to do more than a typical choropleth map.Įven the choropleth was problematic, the authors noted. It noted that the shapes of the city’s boroughs, about which one might want to show comparative data, didn’t actually lend themselves to this application. It discussed a new way to display data about London. What is a Tile Grid Map?Ī bit over a year ago, I ran across an article titled, London Squared Map – Making the City Easier to Read. of Education program that provides the most comprehensive picture of students’ academic progress and performance (Forum One works with NAEP as a subcontractor to the Educational Testing Services (ETS).
#JUSTINMIND TILE GRID SERIES#
To that end, in this latest in my series of reviews of good data visualization practices, I consider an increasingly-popular mapping approach: tile grid maps.Īs a reminder, this series grew out of our work with the National Assessment of Educational Progress (NAEP), the U.S. As I consider data visualization practices, I would be remiss not to include ways to display map data, as it is clearly a prevalent data visualization case.
0 kommentar(er)
