Map Color Breaks

Methods for creating color ramp breaks

When creating a choropleth map, there are various methods available to create color ramp breaks; that is, to split the data into groups and set numeric bounds of each one. The resulting "buckets" are then assigned a color from the ramp.

1. Natural breaks split data into groups such that the variance within each group is minimized, and the variance between each group is maximized. In mathematical terms, the sum of the differences between the grouped values is always the same. This is generally complicated to do manually, and is thus typically determined algorithmically.

2. Equal interval splits the data according to equal numeric ranges. For example, if the data values range from 0 to 100, and we want four equal intervals, we'll bucket the data into the ranges of 0-25, 26-50, 51-75, and 76-100.

3. Logarithmic methods display data with very large ranges in a compact way, such that very high outliers do not skew the color ramp too strongly.

4. Quantiles split data into groups of equal size. For example, if there are 20 data values total, and we want four quantiles (a.k.a. quartiles), we'll bucket the data into four groups containing five values each.

These four methods are shown below:

Of the maps displayed above, the logarithmic color distribution (bottom left) is the least appropriate method to display the dataset.

Although the logarithmic method shown above does not display the dataset well, keep in mind that this is not always the case – different methods work better depending on the particular distribution of data. For example, the logarithmic method is most useful when dealing with very wide ranges of data.

To learn more about how to choose color breaks based on your data, here's a great article from Lisa Charlotte Muth.