The curious case of BrightBands ColorData

Ideas
1
Open in Wolfram Cloud ↗
1 Like
2

Is this really where the divisions are? How did you extract that data?

3

Hi Adam,
I was perusing the ColorData schemes for a project and stumbled onto the partitions of these hues not being located at 6ths. So I decided to make an illustration of it and succeeded with some help from Hans Milton in this thread . To find the discontinuities I employed a telescoping algorithm to 1st determine the numerical resolution of the data, and then telescoped onto the locations. I’ve no idea why the data are this way - perhaps just some concatenations or generated from a different color coordinate system. There is at least one more like it in that collection.

4

Very neat! I thoroughly enjoyed writing this code

changeover[test_,start_,step_]:={start,step}//.{x_,y_}:>If[test[x+y],{x+y,y},{x,y/2.}]
exhibitchanges[listofchangesepsilons_]:=(ColorData["BrightBands"]@#&/@
    {First@#,First@#+$MachineEpsilon})&/@listofchangesepsilons
Table[changeover[
    EuclideanDistance@Delete[List@@ColorData["BrightBands"]@#&/@{#,i/6+.05},0]>.3&,
i/6-.05,.05],{i,5}]
exhibitchanges@%

which agrees with your values (to within 10^{-18}).

2 Likes
5

Wow, that’s compact. Thank you for the programming insight. :slight_smile: