
Enriching visualizations with color categorization
Color is a strong feature in Spotfire and an important visualization tool, which is often overlooked by report creators. It can be seen as merely a nice-to-have customization. However, color can be the difference between creating a stimulating and intuitive data visualization and an uninspiring and even confusing corporate report. Take some pride and care in the visual aesthetics of your analytics creations!
Let's take a look at the color properties of the table visualization. These are only available in the Analyst clients (however, you can still choose a pre-configured color scheme, or choose colors from the legend of a visualization if you're working in web-based clients):
- Open the table visualization properties and select Colors, and then Add the Total population (Gapminder) column:

- Now, you can add points and color rules, or adjust the color gradient. To adjust the existing color gradient, just click on the box next to the Max point and pick a new color, and do the same with the Min point:

- To add a point with a color indicating the average, click Add Point. Spotfire will add a point with a specific value, so we need to change it to show the average. Click the drop-down to the right of the point and choose Average, then select a color that you'd like use to indicate the average within the gradient. This book is in grayscale, so it's difficult to represent the colors accurately, but I recommend starting off at a deep color for higher values, leading to a lighter shade of the same color for the average. Then, choose a totally different color for the minimum value, but one that does not clash with the gradient from high to average values:

You can also use some built-in color schemes in Spotfire—just click the icon that looks like several sheets of paper to expand the menu:

Out of interest, just look at the difference between the spread of the average and the max and the min and the average. I suspect that China and India really skew the population figures. In fact, we can check this out:
- Close the dialog and return to the main visualization.
- Select all the data in the scatter plot.
- Sort the Total population column by descending. Now, look at the data:

You can see from the values that China and India are indeed responsible for skewing the data—their populations significantly outnumber those of the next two countries—USA and Russia. Note that I was still looking at data for 1974 at this point. Your results will be different if you are looking at a different year.