-
Use descriptive axes and titles, include units, and make the fontsize large (~16 pt.). Include other relevant technical information (like the DFT functional you used) in a textbox below the image.
-
Use appropriate color schemes. Colorbrewer can help you pick combinations of colors which look good. Use different shades to accomodate colorblind users.
-
Don't put too much data on a single plot. What are you trying to say? Like your writing, your images should "tell a story".
Note: Many past QE, Prelim, and other presentations can be found on Box. Use the images here as a reference: What works? What Doesn't?
-
Seaborn is a Python package which can make better, more sophisticated plots than matplotlib
-
Matplotlib Documantation is useful. The tutorials can help a first-time user get started.
-
A wide range of color options are available with the matplotlib 'xkcd' tag (yes, that XKCD), the list of colors is here