Sequential palette evaluations

[mcstone]

Created the following sequential palette: #008696, #0d9797, #42a799, #65b69c, #86c5a1, #a5d4a9, #c4e2b4, #e2f1c2, #ffffd4

• It fails color name discriminability (of course). Arguably, this should only be tested for categorical palettes
• It fails all the CVD tests, which are designed for categorical palettes. But for a sequential palette, it's actually fine. Diverging palettes, however, could fail a CVD test (details on demand)

[mcnuttandrew]

This seems to suggest that there should probably be some type of CVD test for sequential, but different than the one we currently have. Do you have any suggestions?

[mcstone]

I'm not sure if there needs to be a CVD test for sequentials per se, as it's really only the lightness steps that matter. For a diverging palette, however, there should be a test that tells you a red-green palette is a problem, but blue-orange isn't. I don't have an algorithm all worked out, but it's roughly this: First, check the endpoint colors. Then step your way towards the center, checking colors of similar luminance to see if they are distinguishable. This assumes the diverging map has the same number of steps on each side of the middle, which have roughly the same luminance.