Most likely, one view of a dataset is not enough. Different graphical representations may have their own advantages and biases. Each would offer a partial set of insights from the dataset.
It would be good to have a few different representations so that we can view the dataset in complementary ways. This is likely to offer the viewers more insights. We are not talking about slicing or dicing here.
In this Viz of the Day on 7/7/2016, the author Nick Hura created a viz on the rankings of 50 US States in 5 different metrics. The visual looks great and attractive! It tells us the rankings of each state quite effectively. Those little bars on a big square is a great design. The map layout is a popular index view of the states.
What I found hard to see is the comparison between the states. Thus I added a sorted view so that we can see visually where each state is sitting, and which states are above or behind. Click the image to view the interactive version.
The added view is based on bump chart, which we often use to view rankings across time or categories.
The new view is presented using the visual tooltip technique. The key steps in this design are:
- Create the bump chart
- Create a dimension "One" with value 1. This field shows 1 in all the rows.
- Create an action filter based on "One". This will turn on/off the entire bump chart.
- Create a highlight action filter based on State. This will highlight the selected state.
With the added view, we have a better idea where each state compares with others.
No comments:
Post a Comment