2

I have some direction (bearing from north) data in degrees. Some of this data jumps between 1-360 degrees when plotted as a line graph, which is not easy to read, as the variation is actually only 1 degree. What is the best way to graph such data (preferably using Excel)?

Unfortunately, applying a fixed value to all values is not an option as the full range of 0-360 degrees is covered in the dataset.

ijf
  • 21
  • Have you looked into polar coordinates? http://en.wikipedia.org/wiki/Polar_coordinate_system#Converting_between_polar_and_Cartesian_coordinates – RandomUser Apr 22 '14 at 20:24
  • I have actually plotted with polar coords already. I should add that ideally, I'd like to plot bearing over time. – ijf Apr 22 '14 at 21:07
  • If you're not opposed to using them, then you could do bearing as the angle and time as the distance from the centre. Though this assumes your times are all non-negative. If some are negative you could offset them so that the origin is at your smallest time. – RandomUser Apr 22 '14 at 21:18

2 Answers2

2

The best I can think of is polar graph $\theta=f(r)$, where $\theta$ is your angle data and $r$ increases with time. I don't know how to plot such graphs in Excel directly, so I converted into $XY$ coordinates using $x=r\cos \theta$ and $y=r\sin \theta$, then used scatterplot with connecting lines. I think the evolution of angle over time is clear.

polar plot

I removed the distracting Cartesian grid and added a red point to mark the origin.

0

Another option is to assume the change between points is less than $\pm 180^\circ$ and add or subtract multiples of $360^\circ$ to each point to make that true. It means your plotted values may go outside the range $0-360$, but the variation from point to point will be represented correctly. You could give that a try.

Ross Millikan
  • 374,822