INTRODUCTION & INSPIRATION
California is the third-largest by area (163,696 square miles) and one of the most popular places for road trips and in the US. California is famous for its diverse geographical landscapes, mountains, beaches, lakes, national parks, and several volcanoes. A majority of California’s cities are located in either the San Francisco Bay Area or the Los Angeles area. This idea for this lab was to create a map of the main California roads based on the dataset of nodes and the roads connecting them. My research question was inspired by the idea to explore the geography of the state and learn more about possible travel routes. I’ve started my research with two articles about California road trips: 10 Super-Scenic California Road Trips and 5 California Road Trips For a Quick Weekend or Longer Adventure. As a source of inspiration for network design, I also used the “A Network Map of the World’s Air Traffic Connections” and the “Where the world flies: Watch the mesmerising maps that reveal the most popular air travel hubs around the globe” articles to create bright roads on a black background.
PROCESS & MATERIALS
In the lab, I aimed to create a map of California state based on road network information. Gephi open source software was used as a tool to visualize the data and create the network. A high amount of data was processed to create the network. I explored the California road network dataset which included a total of 1965206 nodes and 5533214 edges. Intersections and endpoints are represented by nodes and the roads connecting these intersections or road endpoints are represented by undirected edges. I downloaded a .txt file from Stanford Large Network Dataset Collection, converted it into a CSV file, and imported it into Gephi. I’ve compared the performance of Expansion, ForceAtlas, and ForceAtlas 2 to find the best option. ForceAtlas 2 layout showed the clearest result and was used to create the final version of the network. The range of edges was used as a filter to adjust the view, create a clearer image, and visualize possible roads. I’ve also run modularity and used different colors to make the roads distinguishable.
RESULTS
As a result, I got the network of intersections or road endpoints which were represented by undirected edges.
To analyze the result I juxtaposed it with the map with the main roads of California. The pattern of roads looks similar to the actual map, but I’m not 100% sure about the accuracy of the map because of the coastline (left part) which is supposed to be empty of any information about the road network.
REFLECTION
was data processing time, it took about 30 minutes to create a network and 1-5 minutes to process each modification. Gephi also got crushed from time to time. It might be caused by high data volume and my laptop capability. For further work it would be interesting to analyze the road network within one city (which I know well) and see how clear the map is and whether it’s possible to place main sights, government institutions, schools, etc. based on the network. I’m also interested to experiment more with processing time and compare, for example, 10-minutes and 1-hour results.
SOURCES:
Gephi. The Open Graph Viz Platform, https://gephi.org/
California road network, https://snap.stanford.edu/data/roadNet-CA.html
5 California Road Trips For a Quick Weekend or Longer Adventure, https://www.mydomaine.com/california-road-trip-ideas
10 Super-Scenic California Road Trips, https://www.visitcalifornia.com/experience/10-super-scenic-california-road-trips/
Where the world flies: Watch the mesmerising maps that reveal the most popular air travel hubs around the globe, https://www.dailymail.co.uk/sciencetech/article-3618871/Where-world-flies-Incredible-maps-reveals-popular-air-travel-routes-globe.html#ixzz4AKkzWW6p
A Network Map of the World’s Air Traffic Connections,https://www.visualcapitalist.com/air-traffic-network-map/
Stanford Large Network Dataset Collection, https://snap.stanford.edu/data