Originally, I had high hopes of being able to map out and discover what some of the other societal changes you could find in France given the recent shift in favor for certain political parties. I was hoping to find cultural data based over a time period that would mirror the last several election cycles. My modus operandi for information visualizations however, has become somewhat of a mess at the starting points. This vague data set proved too wide of a viewpoint and I soon was just scrounging for any relevant data at all. Certain datasets I used weren’t being properly uploaded into CartoDB either and I began to panic a bit at the lack of progress. Even the relatively straightforward election data proved trickier than planned, mostly because France has so many political parties, but also as a result of less than basic data formatting. Ultimately, I couldn’t apply a lot of what I had. Data issues aside, I still hope to reform the election data and put it to use in another project.
Eventually though, I found the website Opendata Paris. It contains datasets produced by various government-run organizations throughout Paris and researches have the choice to find data based on keywords, themes, producer, and type of file. This last option proved essential since I was hoping to skip some of the data problems I had earlier and restrict my data to geographic files only.
On Opendata Paris, I found a map of all the cafés throughout the city serving coffee for 1€. I also found the geographic files for all of Paris’ museums and its condom vending machines locations. At the risk of proclaiming a stereotype, I did think these locations somehow reflected the expectation of Paris to be a city of café culture and, well, romance. This set me on a mission to find something less Parisian – at least in theory. So, I started looking for a dataset for one of the most archetype American exports – McDonald’s. After my earlier data hunting struggles, I was surprised to find that GoogleEarth made this information really easy to find and download.
Still, layering all of these location types onto the map of Paris, my first thought was that no patterns at all were to be found. Each location type seems evenly (impressively so) spread out throughout the city. So I then had the idea that, rather than distribution, I should start by looking at actual amount of each location type. The hierarchy here was more easily definable. The largest number of locations by a large margin was the category of cafés serving coffee for 1€, followed by museums, then McDonald’s, and eventually condom distribution machines. Once I had established the density rankings of each location category, I could start analyzing this data one layer at a time, then two layers at a time and so forth. Clicking the image below brings you to a map where you can use interact with the locations map as well.
There appears to exist an almost anti-relationship between museums and condom distribution – with the location frequency of condom vending machines leaning to the east perimeter, whereas museums are typically central and into the west of Paris.
In the end, despite its addition to the data as a non-Parisian icon, it seems McDonald’s is in fact the most evenly distributed of these location types. Its franchises are in every arrondissement, even the fancier ones. Even the category with the greatest number of locations, Coffee for 1€, is rarer in the 16th arrondissement – notably the wealthiest neighborhood.
Eventually, I started to seek out a spot in Paris where you could easily attain the benefits of every location type; museum space, cheap coffee, protection, and maybe some fries. I can’t say I think the Lenin Museum (in the 14th arrondissement) would be a particularly romantic (cheap) date night, but perhaps the tropical aquarium (12th) or the Balzac Museum (16th) would make for better choices.
Moving forward with this visualization, I’d like to add to more of a sliding scale of layers – one that moves through the least dense location type to the most (e.g., Condom vending machines to 1€ coffees). This might be something that is already available, limited only by my skill level. It is almost represented now in the map by selecting layers to make visible using a check box method.
More importantly, I would like to incorporate wealth data into the map. While I can vaguely say that the 16th and 8th arrondissements of Paris are the wealthiest in the city, pick out certain bluer collar neighborhoods, and highlight the main tourist drag, it would obviously be better to show this information visually rather than express it anecdotally. I found a solid dataset through INSEE (Institut national de la statistique et des études économiques / National Institute for Statistic and Economics Studies) but I haven’t been able to successfully make that data geographically readable through CartoDB.
Essentially, I have a general hypothesis that the economic gaps in American cities are more drastic than in Europe, and that social elements are likewise a mirror of that distribution. To best visualize this, I’d like side-by-side maps of Paris and New York (or Chicago) with neighborhoods reflecting wealth range and the societal elements layered over each to be filtered simultaneously.
After hours of research and data sourcing and general futzing around with the data and the programs, I found an amazing visualization done by a group of students out of HETIC in France. It was a somewhat tragic ending to this project to see a better version of it, with more information included, already created but maybe I’ll turn that into inspiration.