Upcoming webinar on 'Inforiver Charts : The fastest way to deliver stories in Power BI', Aug 29th , Monday, 10.30 AM CST.    Register Now

Demystifying Radar & Polar Charts

Radar and polar charts use polar coordinates instead of our usual (x,y)-system to create unusual and engaging charts that hold the reader’s attention. Click here for an introduction to radar charts and polar charts. In this article, we discuss the advantages that these charts offer, as well as some drawbacks. We conclude by exploring some best practices that you can adopt to help you make beautiful charts that faithfully represent your data.

Advantages of radar and polar charts

Radar and polar charts offer two major advantages:

1. Compactness

Radar charts are an excellent tool to pack a lot of information into a small area. Consider the parallel coordinates plot, the clustered bar chart, and the radar chart below, all three of which plot the same data. Notice how the radar chart occupies far less space than either of the other two charts. This is especially useful for multivariate data that is tracked over several series, as well as for dashboards with limited space, where the density of information on a radar chart can become a major advantage.

parallel coordinates plot 8 var
clustered bar 8 variables
radar chart 8 variables

2. Novelty and engagement

Both radar and polar charts are relatively unfamiliar representations to most readers, but this novelty can also draw attention to the chart, inviting the reader to investigate further. This increase in engagement can be vital in a crowded media environment.

Disadvantages of radar and polar charts

Along with the advantages discussed above, radar and polar charts also present many difficulties.

A) Perceptual challenges

Our visual perception allows us to most accurately estimate distances measured from a common baseline, as in a line chart. In radar and polar charts, however, we need to estimate radial distances, which can be much more difficult to do. Even with gridlines added, it may be difficult to compare values across a single series. In addition to this, we also need to estimate angles for polar charts, which can make reading values doubly difficult.

B) Overrepresentation of larger numbers

As discussed before, radar charts use polygons to represent multivariate series. However, this means that larger values tend to be overrepresented as the area of a polygon grows geometrically, while our axes typically have linear scales. For example, the following chart shows two coffees scored from 1 to 5 on three flavor metrics. Coffee A scores twice as much as Coffee B on all three metrics, but the blue triangle representing Coffee A is four times as large as the orange triangle representing Coffee B!

Coffee A vs. Coffee B

C) Unfamiliarity of polar coordinates

Polar coordinates can be unfamiliar and confusing to many readers, making the interpretation of charts that use these coordinates difficult. Radar charts may be easier to understand than polar charts in this respect because they can be interpreted as a radial form of the parallel coordinates plot as discussed, but even radar charts often require some introduction.

D) Spurious relationships

With multivariate observations, it is important to pay attention to how we group observations. When we connect the points on the axes in a radar chart, we may suggest relationships that don’t necessarily exist between neighboring variables. Changing the order of the variables may change which relationships are implied, along with the overall shape of the chart, making the relationships themselves arbitrary.

Best Practices

1. Limit the number of series plotted

It is advisable to limit the number of series plotted to allow the reader to clearly see all the series and compare their values. Adding too many series can lead to an occluded or crowded chart.

2. Limit the number of dimensions plotted

It is also best to limit the number of axes to avoid an overload of information, and to allow readers to accurately judge values.

A potential exception to this rule is with sequential, rather than categorical variables. The chart below shows pizza sales for every hour that a 24-hour pizza parlor is open. This is an example of a sequential variable, where we have 24 radial axes plotted (one for every hour), but the main message of the chart remains clear: there are three spikes in pizza sales over the course of the day – during lunch, dinner, and late at night.


3. Introduce your chart appropriately

Radar and polar charts are unfamiliar to most audiences and benefit from an introduction on what they represent and how to read them.

4. Choose the variables and the order of the axes carefully

As with all multivariate data, it is important to choose variables that have a logical connection to each other in order to make meaningful comparisons. In addition, choose a meaningful order for the variables and order the axes of your chart accordingly.

5. Normalize the axes

When comparing different variables, we often have different units of measurement and different scales used for each variable. This means that each axis needs to be normalized for radar charts to fit the correct range of values to each axis. To understand this, let us take the example of a student’s performance at school. 

Jake’s performance at school is measured across four metrics – his mid-term test scores, his final grade, his attendance and the number of assignments he hands in. Now, his mid-term test is scored on a scale of 0 to 50, out of which he scored 40, his final grade is a letter grade, going from A+ to F-, with Jake scoring a B, his attendance is scored as a percentage and Jake attended 80% of his classes, and he had 20 assignments in total to hand in over the course of the year, out of which he handed in 18 successfully. Consider the radar chart below without normalized axes, where all axes range from 0 to 50. Of course, this is not possible for the final grade, which is a letter, but notice how skewed the values appear – Jake appears to have barely attended any classes and seems to have handed in less than half of his assignments. 

school performance wrong

However, when we normalize the axes, setting the ranges and scale individually for each axis, we see that he has performed reasonably well.

school performance correct

This principle extends to polar charts - we must also ensure that both the radial axis and the values represented by the angle also have the correct ranges that correspond to our variables, as in the chart below. Here, the radial axes range from 0 to 400,000 units, and the angle from 0 to $3 million.

polar revenue vs quantity sold

6. Use for quick overviews

Radar and polar charts are better suited to providing an overview of the broad patterns in the data rather than accurate evaluations of the values, given the challenges with estimating radial and angular distances that we discussed earlier.

7. Use transparency

When shading in the area enclosed by the polygons in a radar chart, use transparency to allow gridlines as well as other series to be seen easily.

coffee flavour profile area radar

Using these best practices, you may leverage the power of radar and polar charts to your advantage. Click here for a free copy of xViz to try out radar and polar charts for yourself in Power BI!

-By Hamsini Sukumar

Other resources:

Radar charts 101: Overview and use cases

Polar charts 101: An in-depth guide

Radar vs. Polar charts – A comparison

Inforiver Logo
IBCS Certified
Power BI certified
About Inforiver!

Inforiver drives business productivity and business performance with faster time to insights inside Power BI through an intuitive no-code experience. The product is developed by Lumel Technologies (formerly Visual BI), which has a decade of experience in building add-on solutions on top of BI platforms (SAP BI and Power BI) with our suite of innovative products such as ValQ, xViz, and BI Hub
5920 Windhaven Pkwy, #130
Plano TX 75093
BI Hub
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram