I wish to simultaneously scatterplot two distributions on the same plot, so that I can see at a glance each distribution, as well as the relationship between them.
https://matplotlib.org/3.1.0/tutorials/colors/colormaps.html shows:
... so if I take a chunk out of Viridis and a chunk out of Plasma, (e.g. using how to extract a subset of a colormap as a new colormap in matplotlib?) I should be good to go.
But then I'm losing the full dynamic range.
Is there any "hack" to restore this dynamic range?
The full "mathematically aesthetic" solution may be to dig into the generating code for the colormaps and regenerate from scratch, but I suspect this is a deep dive.
How do you expect to take a subset of a colormap but still have the full dynamic range of the colormap? That is not how colormaps work.
One way you can solve this is by simply using two colormaps that look vastly different. My CMasher package provides a large set of scientific colormaps, which were all designed to be perceptually uniform sequential and unique in appearance. You can easily find two colormaps that are very different there.