Introduction
My intention during this lab exercise was to see how hotels and landmarks are present together in New York City, and to find out how these two kinds of establishments could be represented in a map. I expected most hotels to be in the Manhattan area, but wondered if landmarks were concentrated in the same way.
Inspiration
Since my lab work would focus on NYC, I sought out NYC maps. One of my favorites shows the location of Twitter and Flickr use. The color scheme is straightforward, with orange representing the upload of photos to Flickr and blue representing Twitter messages. White indicates the use of both services at the same time. The map takes advantage of our mental picture of a city, as the dots take on the appearance of city lights, and the brightness of the white dots suggests heightened activity.
The second NYC map is the result of a community mapping project and, while it is not particularly pretty, it is highly functional. NYrestroom.com provides the location of publicly accessible bathrooms in NYC. A simple filter allows you to search by the type of establishment (book store, coffee shop, etc.) and you can manually manipulate the map or let it locate a restroom close to your current location.
Source: http://m3.mappler.net/nyrestroom/
My final map displays the filming locations of movies shot in NYC. The map easily shows a concentration of films being shot in midtown and lower Manhattan (interestingly, it also omits the Bronx). The little character icons infuse the map with personality and make it more memorable.
Source: http://blog.hotelscombined.com/2015/01/movie-maps-where-films-are-shot-worldwide/
Materials
I located a shape file of NYC landmarks, made available by the Landmarks Preservation Commission. A dataset of all NYC hotels proved difficult to find, but I found a list of places to stay as put together by I Love NY. Since this is a statewide initiative, I filtered the list to include only NYC hotels, which left 268 records. I used these two files to create maps in CartoDB, a free online tool.
Methods and Discussion
I first uploaded my shape file into CartoDB to experiment with a map of NYC landmarks. CartoDB suggested 14 different maps. I tried several of these, which included maps of landmarks by status and borough, before deciding on a Category map by landmark type (Historic District, Individual Landmark, Interior Landmark, N/A, and Scenic Landmark).
In InfoWindow, I added several fields to both the Click and Hover tabs, and changed the label titles to cleaner titles (Borough, Landmark, Landmark Type, Address, and Landmark Status).
I very much wanted to include photographs in some way, so I went in search of a feature that would allow me to do so. When I enabled the type of Click display as ‘image header’, CartoDB provided me with directions on how to display a photo as well as text fields when a location is selected. I added a column in Data View called ‘images’ and added URLs for pictures of a few landmarks. Then I dragged the Images field to the top of the fields list in InfoWindow. At this point, when I clicked on Prospect Park, the window seen below popped up.
I tried changing the map to a Cluster type in the Map Layer Wizard, but there were too many clusters to distinguish between them. I also experimented with Change Basemap (including Antique and Toner) and settled on Dark Matter because it made the blue markers stand out (as in the Twitter and Flickr use map seen above).
Next, I uploaded the dataset (CSV file) of hotel locations and added it as a layer to my existing map. In Map Layer Wizard, I tried a Heat Map. Unsurprisingly, most hotels were concentrated in Manhattan – by zooming in, it was apparent that they are particularly present along 5th and Lexington avenues, though there is also a second concentration in Queens, likely adjacent to LaGuardia Airport. Staten Island boasts many landmarks, but very few hotels.
In both the hotel and landmark layers, I changed the map type to Simple and colored the landmarks purple. After using Add Element to add a map title and the sources of my datasets, I investigated changing the hotel markers to something that would distinguish them from the landmark markers and relate to a hotel as the icons in the movie map above relate to the films they represent. I used Marker Fill to indicate hotel locations with a symbol instead of a dot. Because the markers represented hotels, I used a building with several stories (a ‘flat’ symbol) and changed the marker color to green.
Finally, I added a map legend, indicating purple dots for landmarks and flats for hotels.
Future Directions
The green color of the hotel markers is not as easy to see as I would like. Since CartoDB allows you to create custom marker symbols, I would experiment with creating something a bit bolder. In future, it would also be useful to use a dataset that includes all NYC hotels, not just the ones recommended by I Love NY. Ideally, I would find photos for every landmark in the dataset, though unfortunately it seems that images are only enabled for Click, not in Hover. It would also be interesting to group the landmarks by a measure of popularity, for instance, number of visitors per year. This would be a better indicator of where hotels are likely to be concentrated, as the mere presence of a landmark does not necessarily indicate that people are visiting it. Another way to group landmarks would be to focus on a particular category, i.e., Interior Landmarks, and establish finer groupings, such as homes, theaters, and libraries.