This week’s lab had our class taking on data visualization using a variety of tools provided by services online, namely RAWGraphs and (if you were so inclined) Flourish. The dataset we were tasked with representing was a .CSV file containing several fields pertaining to the popularity of certain Baby names in New Zealand throughout the early 2000s. This dataset, it turns out, is not compatible with a majority of readily available graphing formats available on the provided sites.
While working on displaying this dataset in a (somewhat) digestible format, I worked my way through a couple of tools. The most aesthetically pleasing graph was the above Voronoi Tree graph. The University of Bristol describes a Voronoi Tree Map as follows: “This type of diagram is created by scattering points at random on a Euclidean plane. The plane is then divided up into tessellating polygons, known as cells, one around each point, consisting of the region of the plane nearer to that point than any other.” (Source provided here) The individual regions of the above graph, despite this chaotic and seemingly random placement, actually have meaning behind them. The main criteria I chose to have as a physical element of each region is size—which is directly correlated with the number of children under the contained first name.
Believe it or not, the above graph is actually much more legible than the default option, which was a triangle. The main problem with the triangle was that too many regions were rendered unreadable by compression at the edges of the polygon. Though the above octagon is, admittedly, not much better, it stands worlds ahead of the original in terms of legibility.
The key issue with this dataset and the provided formats is that there’s not enough variation between certain elements within different years. What I mean by this is the phenomenon on the above graph of combined cells, where two names from different years are deemed so similar as to share the same region. That would be the main element I would change; find some way to differentiate cells that are completely different in name and year, but counted as the same due to similar counts.
The power of data visualization is incredibly potent. The ability to portray numerical or otherwise indecipherable information in vast quantities visually allows complex topics to reach a wider audience. I’d much rather look at the above graphic than pore over a .CSV file to figure out that the names Daniel and Benjamin had similar headcounts in 2007 and 2004 (respectively). Lin’s talk with us really reinforced to me that seeing is believing, and if we want people to truly understand the magnitude of our world, data visualization is a good place to start.
— A.J.
3 thoughts on “Week IV: Lab IV—Making Graphs”
Leave a Reply to Maddy Brown Cancel reply
This site uses Akismet to reduce spam. Learn how your comment data is processed.
I really like your graph! I tried to make a Voronoi Tree Map as well but I struggled to make it legible too. Every time I tried making the names less “clunky” and more spaced out, it would in turn make the names harder to read for the viewer. I think you did a great job at finding that balance of both legible and well-spaced names.
I really like your graph! My only edit is that the names stack on each other although as someone who played around with the data myself, I know it was a hard issue to combat. I do think the graph you chose was cool although I do think it is hard to get a sense of the data based on the graph. If you had made one for each year it might have been easier to comprehend.
Cool post! Your Voronoi Tree graph is really interesting and the algorithm behind how the tiles are created and placed is neat, too. I had the same issue you did when I was making my data visualization of having the data be unreadable because of compression and density in a limited space, and I think you did a nice job of recognizing that and altering your graph to make it readable and clear to the viewer. I also really like the colors of the tiles in your graph!