jpcm 0.1.28

A perceptually uniform colormap generator equipped with a traditional-ish Japanese color palette. Allows unique colormap generation and integration with ML experiment workflows.

perceptual-jp-colormaps

A perceptually uniform colormap generator for Matplotlib equipped with traditional-ish Japanese colors to serve as a fixed color palette.
Additions are welcome.

Why perceptually uniform colormaps? Can you not just use default Matplotlib colormaps?

• non-perceptually uniform colormaps induce dangerous artifacts, as seen in the below image: . The left colormap introduces new features to the data.
• Matplotlib perceptually uniform colormaps are simply not visually appealing and sometimes lack enough contrast. With this package, anyone can design colormaps to fit their visual style.

dependencies:

• Python3:
  -- colour-science
  -- numpy -- logging
  -- scipy
  -- json
  -- pynverse
  -- matplotlib

tested on:

• Windows 10
• Linux
• Python 3.8, 3.10, 3.12

expected to work on:

• MacOS
• all Python >=3.6 (support available for >=3.8)

install:

• pip3 install jpcm or download source

run generator:

• to create your own colormaps
  • as a package:
    • import jpcm
    • jpcm.register() will register all default colormaps and any additional ones via the optional custom_maps argument. Note any custom_maps should be of the following format (note the RGB key colors):
      cmaps = { 'def' : [[0,0,0],[255,0,0]], }
    • The optional datafile argument will cause jpcm to save all colormaps to that location as well.
  • directly via code:
    • edit maps.py with your preferred cmap styles, and run python3 gen.py or python3 gen.py --readme from /gen/

load colormaps:

• import jpcm
• jpcm.open(<filename>) will return a dictionary containing all the colormaps in the file at .
• Note the file should be of the same type as the generated 'cmap.txt' from the generator

gallery