sujeff@wustl.edu, ryanmiller@wustl.edu
Link to GitHub Repo Link to GitHub PagesWe were inspired by the visuallization shown in class that displayed NYC subway traffic in real time. We think that their use of the map and dot overlay provides easy-to-reference visuals for users on the go. Additionlly, we've seen quite a few maps across the internet that make great use of regional overlays, cool tooltips, and more.
Initially, we considered a dotted map similar to examples that we saw in past class projects. We utlimately realized that shaped visualizations of walkability might not be very useful at a glance, instead, we opted to color-in the neighborhoods of interest and show the average walkability of each neighborhood on the map. Adding assets on-top of the map can be a great way to create an interesting visualization that users will want to explore in in article setting, but are less useful when we consider a user trying to glean information from the map.
Like we mentioned above, we explored (in Figma and in sketches) several different ways of displaying the walk
/bike/ride scores. One such option was adding in bubbles on top of the map, sized based on the walkability score
of the region. We decided that while this method was visually interesting, it wasn't very informative as it is
difficult to compare the sizes of the bubbles and then mentally relate those to scores. We also considered removing the map
completely and just showing the scores of each region. While this is pretty clear if you know what neighborhood you are looking
for, it removes the convenience of a map and the ability of the user to explore.
Ultimately, we settled on shading the neighborhoods based on score, akin to some
sort of heatmap, and then showing their exact walk/bike/ride scores in a tooltip
on hover. This maintains highly available information for users, as it is
easily scannable and easy to understand with the help of a legend. This ensures that users
won't have to guess what the colors mean and their relevance to one another. Because this is
colored according to score, it is easy to watch the gradients change as you move across the map
and view the relations between neighborhoods.
The most important visualization that we implemented is the base map that shows the different neighborhoods and are shaded according to their walkability scores. On this map, the user can view a legend that shows the different colors and their meanings, a toggle that allows for the user to switch between the different walk, bike, and ride scores, and tooltips that show on the hover of a neighborhood. We've also included a legend that maps the color to a description of the score. Additionally, clicking on the walk/bike buttons will change the map color to show ONLY the walk or bike scores (depending on which button is selected).
Through the implementation of this visualization, we learned a lot about the complexity of maps and the many layers that are in play in the most complex of maps, such as Google Earth. We also learned how to mess with location data and coordinates to create useful insights. We think that we answered some of our questions pretty well, but would need to delve much deeper into the many factors impacting the walk/bike scores to come to a better conclusion. We could improve our visualization by adding more detail into each neighborhood, such as which routes are most optimal and have the best walk/bike scores. Additionally, it would be interesting to create some sort of model that can write descriptions for each area to give non-residents more qualitative information about walkability.
From living in St. Louis for a few years, we have noticed that access to public transportation and walkability is widely variable throughout St. Louis. This is a serious problem, as the car-dependent infrastructure which has become the norm since the urban renewal movement of the 50s and 60s is unsustainable, environmentally harmful, deadly, and contributes to segregation. Further, the negative effects of car-dependent development and the resulting lack of walkability has an overwhelmingly disproportionate effect on racial/ethnic minorities and impoverished people. We think that visualizing walkability, bike ability, and transit access in St. Louis in a novel way could provide a greater understanding of this issue, and illuminate areas where an increased emphasis on green, accessible infrastructure is necessary.
Users/viewers of our visualization might be wondering a few different things:
These questions, and many others that pedestrians/cyclists may have about St. Louis, will be the focus of our visualization. If we can answer these questions, the benefits are:
We plan to use the APIs offered by walkscore.com, which includes APIs for walkability score, bikability score, and transit score. These APIs can provide these scores given a location. We also may use data provided by the St. Louis Metro ArcGIS Portal to overlay information about St. Louis' public transit system.
We do not expect that the data will require substantial cleanup, as we expect that the data from the sources mentioned above will be in a fairly clean format and should have complete information for an urban area. We plan to get geographic information about St. Louis' public transit network from the St. Louis Metro ArcGIS Portal. We plan to use the scores provided by walkscore.com to compare the walkability, bikeability, and transit scores of different areas. We expect that we will have to process the data to some degree so that we can display it using a map with d3, potentially with d3-geo. The ArcGIS data should be geojson, so hopefully d3-geo will be able to handle it without much processing.
We think that users must be able to choose a certain neighborhood/area in St. Louis and get information on its walk/bikeability and general accessibility. Additionally, users should be able to view some sort of map of the city (not necessarily the whole city, or a 1:1 map, but some sort of map vizualization).
It would be interesting to allow users to create a pathway, or define some sort of walk path, and then allow users to view statistics for that specific path. Also, it could be interesting to allow users to fine tune their definition of walkability/bikeability based on parameters that we set, so that users with different levels of comfort and ability can granularize their routes and outcomes.