Nicopal

Context

Nicopal provides perceptually uniform and colour-vision-deficiency-friendly colour palettes for scientific data visualisation. Inspired by the work of Fabio Crameri [1], these palettes are designed to represent data as faithfully as possible, ensuring consistent interpretation across all observers, regardless of their colour vision. Their design follows the perceptual uniformity guidelines of Peter Kovesi [2] and uses the colour-vision deficiency simulation model of Machado et al. [3]. They are also part of the broader effort, initiated as early as 2004 by Light & Bartlein [4], to move away from rainbow-type colourmaps, which are known to introduce perceptual artefacts and mislead interpretation.

These palettes can be used as a drop-in complement to the colour-vision deficiencies section of Matplotlib. For more context on colourmap best practices, see the colourmap selection flowchart and the Matplotlib colormaps documentation.

Developed by a protanope (red colour deficiency) student, for everyone.

I. Installation

In a Python console or terminal :

pip install nicopal # Standard installation
pip install nicopal'[advanced]' # Advanced installation

The advanced installation includes scipy and cartopy, which are only required for pal_demo_advanced (see Section 9).

II. Usage

0. Loading the package

import nicopal as ncp

1. Display the package version

print(ncp.version)

Displays the version and some information about the package.

2. Quickly test all palettes

ncp.pal_test()

Checks that all palettes load correctly and displays a confirmation message if everything is valid.

3. List available palettes

ncp.pal_list()

Returns a sorted list of all available palette names.

4. Visualize a palette

ncp.pal_show("Lithium")

Displays a gradient of the chosen palette. pal_show can be used with any palette from pal_list().

5. Visualize all palettes

ncp.pal_all()

Displays all palettes and their associated names (see Section 11).

6. Load a palette as a list of HEX colours

ncp.pal_hex("Lithium") # all stored colours
ncp.pal_hex("Lithium", N=10) # resample to exactly 10 colours

Returns the palette as a list of HEX color codes (e.g. ['#1a2b3c', ...]). See the Matplotlib colour formats documentation for details on how to use these strings.

7. Use a palette as a colormap

colormap = ncp.pal("Lithium") # normal order
colormap_r = ncp.pal("Lithium", reverse=True) # reversed via parameter
colormap_r = ncp.pal("Lithium_r") # reversed via '_r' suffix
colormap_r = ncp.pal("Lithium_r").reversed() # reversed via Matplotlib native method

ax.contourf(x, y, z, cmap = colormap)

Palette names are case-insensitive: "lithium", "Lithium" and "LITHIUM" all work.
ax.contourf can be replaced by any other Matplotlib function that accepts a colormap.
x, y and z are your data to visualise.

For more information on reversing and manipulating colormaps, see the Matplotlib colormap manipulation guide.

8. Extract colors from a palette

sample = ncp.pal_sample("Carbon", 6)
x = ["A", "B", "C", "D", "E", "F"]
y = [3, 7, 5, 6, 4, 8]
plt.bar(x, y, color = sample)
plt.show()

Returns n discrete colors extracted evenly from the palette.

9. Palette demonstration

ncp.pal_demo("Selenium")
ncp.pal_demo_advanced("Selenium")

Displays various types of plots using the specified color palette. The function pal_demo_advanced requires two additional optional libraries (scipy and cartopy), which enable visualizing the effect of a palette on maps. These two libraries are optional and not required when installing the package; however, they can be installed directly during package installation using the command pip install 'nicopal[advanced]'.

10. Palette score

Display the scientific colourmap validation score of the Nicopal sequential colour palettes.

ncp.pal_score() # For all sequential palettes
ncp.pal_score("Zinc") # For a single palette

The current scoring system is available only for sequential colour palettes.
The methodology, validation, and interpretation of the different metrics used in the colourmap validation score are currently being prepared for publication in a scientific article.

11. Palette names

Boron | Carbon | Cesium | Chlorine | Cobalt | Iodine | Iridium | Iron | Lithium | Magnesium | Manganese | Mercury | Neon | Nickel | Nitrogen | Osmium | Oxygen | Radium | Rubidium | Selenium | Silicon | Silver | Sodium | Sulfur | Uranium | Vanadium | Zinc |

All names are case-insensitive. The _r suffix reverses any palette (e.g. "Lithium_r").

III. Methodology

Documentation in progress...

Bibliography :

[1] Crameri, F., Shephard, G.E. & Heron, P.J. 2020. 'The Misuse of Colour in Science Communication'. Nature Communications 11(1): 5444. doi:10.1038/s41467-020-19160-7. Paper link

[2] Kovesi, P. 2015. 'Good Colour Maps: How to Design Them'. arXiv:1509.03700. Paper link

[3] Machado, G.M., Oliveira, M.M. & Fernandes, L.A.F. 2009. 'A Physiologically-based Model for Simulation of Color Vision Deficiency'. IEEE Transactions on Visualization and Computer Graphics 15(6): 1291–1298. doi:10.1109/TVCG.2009.113. Paper link

[4] Light, A., and P. J.Bartlein (2004), The end of the rainbow? Color schemes for improved data graphics, Eos Trans. AGU, 85(40), 385–391, doi:10.1029/2004EO400002. Paper link

Citation

If you use Nicopal in scientific publications, please cite the package.

Suggested acknowledgement

This work uses the Nicopal colour palettes developed by Nicolas Tharaud.

:) NT