2
$\begingroup$

I want to visualize a 2D sample space imputed from a 2D scatter of measurements. Something like a 2D hexbin color histogram would work if it were enhanced in the following ways:

  • The number of samples in a bin determines the saturation of the color -- corresponding to the confidence obtained with more samples.
  • The hue of the bin is the maximum likelihood or expected value measurement for that bin.
  • Bins with no or very few samples would enjoy higher confidence, hence saturation, based on imputation from nearby bins with higher confidence (number of samples).

This would require a 2D color map with one dimension a range of hues and the other dimension a range of saturations.

Is there published work along these lines, preferrably with an existing software library?

$\endgroup$
1
$\begingroup$

The paper Value-Suppressing Uncertainty Palettes may be what you want. The link is to the github site which links to the paper itself and other materials.

The idea is to suppress values with high uncertainty by desaturating the color. The paper discussing two techniques, with the new technique "VSUP" being to not only desaturate the color but also use a coarser color map for the hues of more uncertain values.

enter image description here

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.