Create Color Maps
This is a Python module to create color maps with pretty much arbitrary size according to the work of Kenneth Moreland as described at
http://www.kennethmoreland.com/color-maps/
These color maps allow mapping a scalar value from a finite range to a color. This is useful in heatmaps or space-filling curves.
Objects with the predefined parameters can be created using the make_* functions
like this:
map = moreland.make_kindlmann()
These objects are for now meant to
- create a SVG file with the color map, usable to show the range
- CSV files with the RGB color values, selectable in saturated (0-255) or real (0-1.0) form, or both
Create Individual Color Map File
Use the get-color-map.py script to get individual CSV files describing the selected color map:
python3 get-color-map.py viridis 256 byte
This will create the saturaged color Viridis map with 256 entries. It is also possible to create the
real version of the map by using float instead byte. Or use both.
The output of the command above is in the file viridis-table-byte-0256.csv which starts/ends with the
following lines:
scalar,RGB_r,RGB_g,RGB_b
0.0,72,0,84
0.00392156862745098,73,0,86
0.00784313725490196,73,0,87
0.011764705882352941,74,0,89
0.01568627450980392,74,1,90
...
0.9882352941176471,236,232,21
0.9921568627450981,238,232,24
0.996078431372549,241,232,26
1.0,243,233,28
The first value is the scalar value, ranging from 0 to 1. By mapping the range [A,B] of the values which
are meant to be represented as color to the range [0,1] one selects the color with the scalar value closest
to the mapped value. The values are equal distance which means that the scalar value in the first column
does not actually have to be used, the index into the table can for a value v can be computed with
(v-A)/(B-A).
The color values are in this case given as a triple of bytes. If the last parameter to the get-color-maps.py
script is float instead of byte the color is given as a triple of floating-point values.
Visualization
The gen-svg.py script can be used to generate a SVG file to visualize color maps. It works only for byte
maps with 256 colors. As an example, the black body map in black-body-table-byte-0256.csv can be visualized
with
./gen-svg.py black-body-table-byte-0256.csv > black-body-256.svg
The result looks like this:
With the additional --fine parameter each diagonal line has the same color. The result is a picture with
a much higher resolution:
Requirements
On Fedora systems not all used Python packages are packaged as of Fedora 33. Explicitly install the packages first with
pip install colormath toyplot