Legends

A legend pairs a plotted symbol or line with a label. Use one when a small number of discrete objects must be distinguished, such as selected grains, boundary classes, or reference directions.

In MTEX, legend entries are opt in. An object appears only when it is plotted with a displayName. This is the opposite of MATLAB’s default behaviour. It prevents markers, construction lines, and a colour-coded background from each acquiring an unhelpful entry.

Choose the legend entries

Name only the objects that the reader needs to distinguish. This example names a point and a great circle. The pole at the centre remains unnamed.

from mtex import *

plot(vector3d.X, upper=True, displayName='point')
hold(True)

circle(vector3d.X, displayName='great circle', lineColor='r', lineWidth=2)

plot(vector3d.Z, marker='p', markerSize=30)
hold(False)

lgd = legend('show')

Read and adjust the result

The figure contains three plotted objects, but the legend has two entries. The central star is visible in the axes but absent from the legend because it has no displayName.

After MTEX has selected the entries, legend hands its options to matplotlib’s legend. Calling it again replaces the legend: an outside location avoids covering the spherical plot, and a title explains what the entries classify.

lgd = legend('show', location='eastoutside', title='Geometry')

Choose the right colour guide

A legend represents discrete identities. When colour represents a numerical value, use a colour bar because it shows the complete value-to-colour scale. When colour represents a direction or an orientation, use a colour key that maps each possible direction or orientation to a colour. The next page introduces numerical colour scales; IPF Maps explains direction colour keys for orientation maps.

The matplotlib Legend properties provide further controls for placement and styling.

References

• S. R. Midway, Principles of Effective Data Visualization, Patterns 1 (2020), 100141, explains how selective legends and consistent visual scales make plots easier to compare.

Next

Continue with Color Maps to control how numerical values are translated into colours.

Technical Details

MATLAB moves and titles the legend by setting lgd.Location and lgd.Title.String on its handle; a matplotlib legend cannot be moved outside its axes after it is made, so the page calls legend a second time with location and title.

@article{bachmann2010mtex,
  author  = {F. Bachmann and R. Hielscher and H. Schaeben},
  title   = {Texture Analysis with MTEX - Free and Open Source Software Toolbox},
  journal = {Solid State Phenomena},
  volume  = {160},
  pages   = {63-68},
  year    = {2010},
  doi     = {10.4028/www.scientific.net/SSP.160.63},
  url     = {https://doi.org/10.4028/www.scientific.net/SSP.160.63}
}