Show description

Let’s use that freedom to indicate the Saffir-Simpson classification for storm intensity at each measurement point. Here’s where we can take advantage of the filter option we included in the data formatting function. The Saffir-Simpson classification is based on wind speed, so we’ll filter based on the wind property. For example, here’s how to extract only those values that represent a Category 1 hurricane, with wind speeds from 74 to 95 miles per hour. The function we pass to get_points returns true only for appropriate wind speeds.

If we need to visualize three variables, we could use a scatter plot framework for two of the variables and then vary the size of the points according to the third variable. The resulting chart is a bubble chart. Using bubble charts effectively requires some caution, though. As we saw earlier with pie charts, humans are not very good at accurately judging the relative areas of nonrectangular shapes, so bubble charts don’t lend themselves to precise comparisons of the bubble size. But if your third variable conveys only the general sense of a quantity rather than an accurate measurement, a bubble chart may be appropriate.

This addition gives us the final chart shown in Figure 1-20. 85 Life Expectancy vs. Health-Care Spending (2010 OECD data) Asia Europe Middle East Pacific Americas United States 80 Years 75 70 5% 10% 15% 20% 25% Spending as Percentage of GDP Figure 1-20: Adding a legend completes the chart. Step 8: Work Around Flotr2 “Bugs” Be sure to refer to Step 9 of “Creating a Basic Bar Chart” on page 14 to see how to work around some “bugs” in the Flotr2 library. info | 33 Adding Magnitudes to X/Y Data with a Bubble Chart Traditional scatter charts, like those described in the previous example, show the relationship between two values: the x-axis and the y-axis.