3 ms·
There are always tradeoffs, but most situations don't require more than just a few colors, and it's still a neglected issue even though it can be solved with ad
by acadapter 5y ago
There are always tradeoffs, but most situations don't require more than just a few colors, and it's still a neglected issue even though it can be solved with adhering to some principles.
https://davidmathlogic.com/colorblind/ https://davidmathlogic.com/colorblind/
https://www.tableau.com/about/blog/examining-data-viz-rules-dont-use-red-green-together https://www.tableau.com/about/blog/examining-data-viz-rules-...
- josephcsible 5y agoI've seen the Wong palette before, but does it actually solve the problem? In particular, aren't #009E73 and #D55E00 quite difficult to distinguish if you have protanopia or deuteranopia?
- acadapter 5y agoI guess it's not perfect, but the point I'm trying to make is that this is an issue that relates to so many people, that there should be some initiative to create an ISO standard or equivalent.
- jryb 5y agoIt is generally called the Okabe-Ito palette since it was created nine years before the Wong paper. But I digress. Colorblind-friendly color palettes are not sufficient in all cases - even a non-color-deficient person will have trouble distinguishing similar colors when the total area is small (as often occurs in thin lines or dots), or if there are simply too many colors in a plot. A good overview on choosing colors when making figures while keeping color-deficiencies in mind: https://www.youtube.com/watch?v=7PsbdY3aWdg https://www.youtube.com/watch?v=7PsbdY3aWdg There really is no perfect solution for all cases, which is why we don't just have a set of hard-and-fast rules for making figures, though there are definitely some choices that are better than others.