Blog Post: Second Iteration (Visualizations) - AlexSanford13/Planes_vs_Animals GitHub Wiki

This second iteration was pretty straightforward. After getting lost down the rabbit hole that was ARCGis online last time, I felt that many of the other visualizations did suffer because of it. Therefore this time I left that visualization alone and focused in on my other information.

First I wanted to update my time of day and month visualizations. After the presentation of my first iteration, I got some great feedback on how to visualize this information. I started with the most obvious, the bar chart. Having had a little experience with Tableau in a previous job, I decided to attempt to create this using those tools. It was fairly straightforward, as it was simply importing the table and dragging these into place. The nice thing about this route was the ability to host and then embed the visualizations.

For the time of day, I had a bit of a struggle. I wanted to do some really cool effect with a sun on a sort of parabolic curve, but could neither find a good one or create my own. So I looked through a variety of tools to see what would do instead. I look at the charts offered on both RAWTools and Tableau. While Tableau was a bit more powerful, it didn't really contain anything nice. Only a bar chart would display the data well, but that was hard to see since the drastic differences. On RAWTools I did find the circle chart that I ended up using. This was nice because of its circular nature I felt not only did display the nature of a day well, but flowed well and properly showed the information.

Next, I had two more visualizations left. I had the damage done, and the top 10 contenders. These two charts were also originally tables that could be far more interesting. I started with the top 10 contenders. This felt like while there was important data in the numbers, the relative size was more important. The box display I decided with in the end displayed that data nicely. It was available on Tableau, so I was easily able to import the data and then embed the visualization.

Finally, I needed to display the damage done. This proved to be a major question. As the one category way outnumbered the other categories. The numbers were just far larger, and most visualization did not look good since all you could see was the one bar, or part of the line. I then stumbled on the bubble chart. While humans do have issues discerning the size difference between circles, it was the easiest to see all the categories while displaying the data. I also included the tool-tips to ease seeing the numbers. I also made the decision to combine the military and civilian data. This was mostly since the no damage category did not differentiate, so keeping the two other military sections separate did not do well.