Does Mathematica have a graphics option that makes plots intelligible for people with colorblindness?
-
2$\begingroup$ Even if one is not colorblind, a plot might still not be intelligible as a result of poor presentation. There are many color schemes available but ultimately I think it is up to you to find a way to present your results clearly; there are no magic options to improve a plot's intelligibility for people with particular visual impairments. $\endgroup$– Oleksandr R.Mar 7, 2013 at 22:32
3 Answers
A nice set of color-blind proof and print friendly color schemes can be find here (see also the pdf). Actually, the distinct color scheme describe in the above links is not very different from the default Mathematica color set.
-
$\begingroup$ Is this the Mathematica default scheme or the scheme recommend for those with color-impaired vision? $\endgroup$ Mar 8, 2013 at 7:34
-
1$\begingroup$ @m_goldberg An answer to your first question: the pictures above are not the Mathematica default color scheme, they are taken from the docs they I linked in my answer. $\endgroup$– VLCMar 8, 2013 at 10:10
-
1$\begingroup$ @jenson I fully agree, data should be distinguished by colors, symbols and line styles to be always easily readable. $\endgroup$– VLCMar 8, 2013 at 11:30
-
3$\begingroup$ I am not fine. I have a diagnosed eye problem -- I can only distinguish color at much higher contrast levels than those with normal color vision. The text I referred to is not "barely readable" to me, it is invisible. The text in the dark red octagon at the right bottom qualifies as "barely readable" to my vision. $\endgroup$ Mar 8, 2013 at 12:59
-
1$\begingroup$ @m_goldberg What I meant is that the point of a color scheme for people with color-impaired vision is to distinguish the colors of the octagons, not to read the text. I've added another version of the same picture, maybe it helps. $\endgroup$– VLCMar 8, 2013 at 13:05
It looks like the new default plot colors in Mathematica 10 take into account colorblindness.
Here's a screengrab from that page.
My answer pertains mainly to color schemes as commonly used in Plot3D or DensityPlot rather than distinct curves in a regular Plot. The short answer is that the following three color schemes tend to be better than most schemes for everyone, including the color deficient: SunsetColors, ThermometerColors, and LightTemperatureMap. The right choice depends on what you're plotting.
I don't know for sure which color schemes are the best for the color deficient, but the site mentioned in @VLC's answer says that sunset is better than a rainbow scheme. Note that the "sunset" on that site looks more like Mathematica's ThermometerColors or LightTemperatureMap.
ColorData["ThermometerColors", "Image"]
ColorData["LightTemperatureMap", "Image"]
Mathematica has several built-in ColorSchemes that you can find in the documentation page guide/ColorSchemes. You can view them all yourself by using the command ColorData and trying them out.
ColorData[]
(*
{"Gradients", "Indexed", "Named", "Physical"}
*)
Within each of these collections, there are several schemes.
Short[ColorData["Gradients"]]
(*
{AlpineColors,<<49>>,WatermelonColors}
*)
Try out a scheme with the following command:
ColorData["SunsetColors", "Image"]
The best choice for you might depend on what you're plotting. For example, if you were showing an astronomical image, SunsetColors might be better because the areas with no flux would be black.
If you were making a geographical map of the Earth, having a blue/red scheme would probably be better.
Note that this site shows that the trench near Japan is more visible to everyone with this blue/red scheme than in a visible spectrum scheme.
-
2$\begingroup$ How is any of this relevant to the question, which was about "color schemes for the colorblind"? $\endgroup$– rm -rf ♦Mar 6, 2014 at 1:15
-
1$\begingroup$ Some schemes are better or worse for the color deficient. As stated in my answer and the site I referenced, these blue/red schemes are better than rainbow schemes for the color deficient. I gave some examples that are consistent with the recommendation of that site, and I included information on how to look at other schemes for those who don't even know where to look. $\endgroup$– PaulMar 6, 2014 at 1:30