A complete guide to area charts

An area chart combines the line chart and bar chart to show how one or more groups’ numeric values change over the progression of a second variable, typically that of time. An area chart is distinguished from a line chart by the addition of shading between lines and a baseline, like in a bar chart.

This area chart shows the number of active users for a fictional web-based company, computed by month. Values for each month can be measured not just from the vertical position of the top of the shape, but also the colored height between the baseline and top. In this chart, we can see that the number of active users has about doubled from November 2019 to February 2020, and that the rate of user gains has increased over time.

While the example above only plots a single line with shaded area, an area chart is typically used with multiple lines to make a comparison between groups (aka series) or to show how a whole is divided into component parts. This leads to two different types of area chart, one for each use case.

Be careful that one series is not always higher than the other, or the plot may become confused with the other type of area chart: the stacked area chart. In those cases, just keeping to the standard line chart will be a better choice.

Generally, when the term ‘area chart’ is used, what is actually implied is the stacked area chart. In the overlapping area chart, each line was shaded from its vertical value to a common baseline. In the stacked area chart, lines are plotted one at a time, with the height of the most recently-plotted group serving as a moving baseline. As such, the fully-stacked height of the topmost line will correspond to the total when summing across all groups.

You will use a stacked area chart when you want to track not only the total value, but also want to understand the breakdown of that total by groups. Comparing the heights of each segment of the curve allows us to get a general idea of how each subgroup compares to the other in their contributions to the total.

Data that is to be depicted with an area chart is typically aggregated into a table with two or more columns. The first column indicates the positions on the horizontal axis where each line will be plotted. Each column that follows will indicate the vertical contribution for each point, one column per series to be plotted. This format is applicable for both the overlapping and stacked area chart, with the main difference between charts the way the values are interpreted for rendering.

For stacked area charts, some visualization tools require the columns to list not the individual contributions, but the cumulative contributions instead. In this case, the columns specify the line heights directly, and the contributions of each group are implied by the difference in values between columns.

While a line chart is not restricted to requiring a zero-baseline, the addition of shading means that the heights of the colored regions will be used to compare the size of each group’s values. Thus, like a bar chart, it is compulsory for there to be a zero-baseline against which the shading should be performed. To cut the axis would result in the actual ratio in group values to not match up with what is implied by the generated plot.

An exception to this rule can come in when we compare two series in an overlapping area chart with a change to the shading rule. If we limit the shading to be between the lines, rather than from both lines to a common baseline, then we can zoom the vertical axis limits into the effect of interest without a baseline. Shading now takes a different meaning, with the color indicating which group as a larger value, and the amount of color the size of the difference.

The more series we have in an overlapping are chart, the more combinations of colors there will be when they overlap. The fact that most colors will not be associated with a single group can cause some difficulty in interpretation. Even with only three series, this can sometimes be too much to track: three individual colors, three pairwise overlaps, and one color for all three groups overlapping totals seven colors in all.

Comparing two series is usually safe, though if one series is always larger than the other, the plot can be easily mistaken for a stacked area plot instead. Readers can also be confused by interpreting the overlapping colors, which will not be present in the general legend. As a general rule of thumb, if you’re thinking about using an overlapping area chart, limit yourself to two series and think about if using a line chart will show the comparison between groups more clearly.

If you want to know about the exact group values and their changes over time, then choosing a standard line chart will be a better option, as demonstrated in this Everyday Analytics article.

A common option for area charts is the percentage, or relative frequency, stacked area chart. Rather than stack the absolute values of each group at each vertical slice, we stack the relative or percentage contribution of each group to the total, so that the overall height is always 100%. This chart type loses information about the trend of the absolute totals (and so will need its own line chart) but helps to bring out the comparison of relative contributions between groups. As a bonus, this chart type gains a second baseline across the top of the plot upon which an individual group’s contributions can be gauged.

The line chart is the main progenitor for the area chart type. If you aren’t sure if an area chart is what you are interested in, using a line chart is unlikely to steer you wrong. This is especially true for the overlapping area chart, where overlapping regions can quickly get out of control. As another consideration, it is always possible to just generate more plots if you are interested in making multiple comparisons in your data rather than just feel the need to choose just a single chart type to carry as much as possible.

As a combination of two of the most common chart types (line and bar), the area plot is also a fairly common chart option across visualization tools. Where the area chart is present as a chart type, the usual implementation is as a stacked area chart. Overlapping area charts are usually not a built-in option for visualization tools, and will need to be custom-created starting with a line chart base. For some tools, often programmatic, creation of either type of area chart is not a basic chart type option, and some additional work will need to be performed on the data to create an area chart from shaded areas and lines.