Values and angles visualization — Part 3

The types of radars presented in the first two articles of this series make it possible, respectively, to display the distribution of angles or to represent non-angular measurements related to an angle measurement. They are, ultimately, representations of a frequency table or table of values, adapted to an angular dimension. A more refined way of visualizing these types of data would be to show the distribution of the non-angular value, itself distributed according to the angular value.
And that’s very timely, because that’s exactly what a radial bar chart is proposing…

The radial bar chart has the property of showing a non-angular measurement in a segmented way, according to an angular measurement. For each “group” of angular measurement, the non-angular measurement is segmented on a regular basis (i.e., its “expression space” is divided into equal parts) and the frequency of occurrence of the values is displayed as a stack of segments of variable size (the higher the frequency, the greater the segment is). Each of these stacks of segments is then displayed in a circular diagram as a sector corresponding to the “group” of angular values it belongs to.
Said like that, it may seem complex, but the readability of this type of diagram is relatively important: a first glance helps to understand the relative importance of each angle sector; next come the segment colors and their relative importance, which helps to understand the frequency of the non-angular measurement.
There is no scale displayed right on the radial bar chart. To keep it legible, this scale is present next to the graph. It is in the form of a sequence of breakdowns.

To set this diagram, you must first select the modules and the measurements you want to display. For the angle, in historical data, you can choose the most appropriate dataset (amplitude, average, maximum, median, middle, minimum or standard deviation, regarding the compilation mode you’ve set). For the measurement you must choose the module and the measurement. For historical data, you must choose the dataset too.
In the angle parameters, you can set the number of sectors on which you want a representation of the values. Allotment is done for 4, 8 or 16 sectors. In the measurement parameters, you can set the number of segments you want to display.

As an illustration, the following chart displays a standard windrose diagram. To do the same, you must choose “wind source” as angle, “wind strength” or “gust strength” as measurement. To refine the visual aspect, just use “international wind standard” as color scheme. That’s it!

Unlike the distribution radar chart and full radar chart, the radial bar chart is commonly used in meteorology… But only to display windrose diagrams. What a pity! While waiting for the fourth and last article of this series, I suggest you to experiment with other measurements types, you will see that one can read some data in a very complete and unusual way.