@Szabolcs does this help at all: mathematica.stackexchange.com/a/64514/38205

@Szabolcs I may be barking up the wrong tree, but here goes: A while ago I dealt with a similar issue when assigning colors to the 6-membered mult. tbl for S3. To brief things down, Josef Albers (en.wikipedia.org/wiki/Josef_Albers) may be of use.

@halirutan Hi man

I want to get the digital position and recognize the digital from a image with c++. Could you give me some direction?

Are we still in a phase where one needs to recreate ticks/frameticks manually only to apply e.g. ScientificForm to labels?

This was ridiculous around V7-V8 but now... And back then one could at least use FullGraphics to postprocess them

@Szabolcs In image processing something similar is used in the farthest point segmentation. It is basically simple: You use a metric color space. Start with a random color and select then a color that is as far away as possible in the color-space. The third color needs to be the one that is the farthest away from both points and so on.

Although Solve method is clearly not optimal, I think it shows how avoiding explicit usage of brute force (exhaustive search) is quite plausible and can make much larger queries easy to implement: mathematica.stackexchange.com/a/161022/3056

I was a bit pessimistic on Solve on this, but I guess some sort of Diophantine equation solver helps a lot in this case.

Sometimes it seems solutions to similar problems are really found only by exhaustive search through all discrete candidates...

Does anyone know of a package that can create quality dendrogram plots? Both HierarchicalClustering and Dendrogram are very limited.

Is there an image processing functions that can increase/decrease saturation, like in an image processing program?

I can convert to a colour space like HSB and scale the second channel. Is there any other way? It doesn't really matter how it's done precisely. It's only for aesthetic purposes.

@Szabolcs Yes, use HSB