[R] palettes for the color-blind

[Thomas Lumley]

On Thu, Nov 3, 2011 at 10:26 AM, Max Kuhn <mxkuhn at gmail.com> wrote:

> First, can anyone verify that these the colors in col2 are
> differentiable to someone who is color blind?
>
> Second, are there any other specific palettes that can be recommended?
> How do the RColorBrewer palettes rate in this respect?

If you go to www.colorbrewer.org, the ColorBrewer site, it has ratings
of the palettes for visibility under a variety of conditions,
including red-green color blindness. Some of them are good, but not
all of them.

The dichromat package attempts to show the impact of both sorts of
red:green anomalous vision on color visibility.  It isn't quite right
because of gamma correction, but people have told me that it is a
fairly good representation, and it does have the right impact on
clustering of pixels in some of the Ishihara color vision tests.   It
suggests that your colors 1 and 3 will be too similar and 2 and 4 will
also be too similar for someone with protanopia.

You aren't going to be able to get five colors that are equal
luminance, equal chroma, and distinguishable to dichromats: you're
putting three constraints on a three-dimensional space and you will
end up with just two points.

For three colors I would suggest orange, blue, gray. More than three
will be hard.

  -thomas

-- 
Thomas Lumley
Professor of Biostatistics
University of Auckland