rang for Python

Color palettes from Persian art. Part of the Rang project, which also packages the palettes for R, ArcGIS Pro, QGIS, GeoLibre and HEC-RAS.

pip install rang

The package has no required dependencies. The matplotlib helpers need matplotlib, which you can pull in at the same time:

pip install "rang[plots]"

Colors

import rang

rang.list_palettes()                      # available names
rang.rang("Kashan")                       # all nine colors, ramp order
rang.rang("Kashan", 4)                    # four well separated colors
rang.rang("Kashan", 30, "continuous")     # interpolated ramp
rang.source("Kashan")                     # the artwork behind the palette

Discrete requests use the stored separation order. Continuous requests interpolate through the source ramp in sRGB.

matplotlib

rang.cmap("Termeh")           # smooth colormap
rang.cmap("Termeh", 6)        # six fixed steps, for classified data
rang.cmap("Termeh", direction=-1)   # reversed

rang.set_palette("Golestan")  # default colors for lines and bars

register() adds every palette to the matplotlib registry under a rang: name. Do this once and any library that accepts a colormap name will take a Rang palette, including xarray, geopandas, rioxarray and seaborn.

rang.register()

plt.imshow(data, cmap="rang:Iwan")
plt.imshow(data, cmap="rang:Iwan_r")      # every palette also registers reversed

Calling register() again is safe and quiet.

Picking a palette

Termeh, Iwan, Khatam and Rostan run steadily from light to dark or dark to light, which is what depth, elevation and rainfall maps need. Kashan, Golestan and Mina move from a warm side through a light center to a cool side. Nasir, Shahnameh and Gilas jump around in brightness, so they suit categories rather than measured quantities.

Each palette page in the repository shows sample plots and the color vision checks.

License

The package software is available under the MIT License. To the extent that copyright or database rights apply, the Rang palette definitions are dedicated to the public domain under CC0 1.0 Universal. Photographs and third-party data are covered by the notices in the main repository.