How to appreciate, use, and choose Scientific Colour Maps
- 1University of Oslo, Centre for Earth Evolution and Dynamics, Department of Geosciences, Oslo, Norway (g.e.shephard@geo.uio.no)
- 2Department of Earth Sciences, Durham University, Durham, United Kingdom
The visual representation of data is at the heart of science. One of the choices faced by the scientist in representing data is the decision regarding colours. However, due to historical usage and default colour palettes on visualisation software, colour maps that distort data through uneven colour gradients are still commonly used today. In fact, the most-used colour map in presentations at the EGU General Assembly in 2018 - including Geodynamics sessions - was the one colour map that is most widely known to distort the data and misguide readers (see https://betterfigures.org/2018/04/16/how-many-rainbows-at-egu-2018/).
Here, we present the work that has been accomplished, the readily available solution, and present a how-to guide to ‚Scientific Colour Maps’ (Crameri 2018, Zenodo; Crameri et al. (In Review)), a methodology that prevents data distortion, offers intuitive colouring, and is accessible for people with colour-vision deficiencies.
Crameri, F. (2018). Scientific colour-maps. Zenodo. http://doi.org/10.5281/zenodo.1243862
Crameri, F., Shephard, G.E. Heron, P.J. Advantage, availability, and application of Scientific Colour Maps. (In Review with Nature Communications)
How to cite: Shephard, G. E., Crameri, F., and Heron, P. J.: How to appreciate, use, and choose Scientific Colour Maps, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11833, https://doi.org/10.5194/egusphere-egu2020-11833, 2020
Comments on the display
AC: Author Comment | CC: Community Comment | Report abuse
cool!
Thanks for the comment! Yes, we think it is cool too, that there is finally a complete set of Scientific colour maps for all plots you can imagine, leaving no more excuses to visually distort data and exclude readers.
Are you able to look into integrating directly into matplotlib? I had a quick look through your site and found some instructions to manually load these colour maps, but in the long term should be made ready-to-use directly through the software-specific channels:
https://matplotlib.org/3.2.1/tutorials/colors/colormaps.html
Thanks!
Yes, that is the ultimate goal to have the Scientific colour maps available as build-in options in all plotting tools and packages. I currently don't have the capabilities to implement them myself into MatPlotLib.
Of course, MatPlotLib has already some perceptually uniform colour maps like 'viridis', but to have a more complete set of Scientific colour maps available (e.g., for more intuitive parameter field representation) would be very useful indeed. Maybe you can help, Ashley? Or else, it is always worth asking the main software developers to implement the suite of Scientific colour maps.
There is already a python package: https://github.com/callumrollo/cmcrameri
That is right, thanks Ales for mentioning it!
The "cmcrameri" package by Callum Rollo is now available through pip and anaconda, and contributions for extentions are welcome via the GitHub link in Ales' comment.
The information about this package with a detailed instruction will be provided in the Scientific colour maps' user guide in a next update.
Thanks for this very important topic. Did not know about your colormaps before but at least i use viridis and similar colormaps.
What do you think about publications with distorted colormaps? Should they be rejected by editors/reviewers? How can we convince journals to make non-distorted colormaps mandatory?
You're very welcome. Yes, I probably don't see all aspects a journal has to take into consideration, so this question should be directed to the individual journals.
However, there is a peer-review process that should take care of problems like data-distorting, reader-excluding scientific studies submitted. It just should finally become clear to every scientist what a colour map really is - it is nothing else than an axis. A direct analogy to a distorting colour map like 'rainbow' is an x-axis with randomly spaced axis ticks (e.g., 1-2 covers a larger distance than 2-3), then it should become clear to any editor and reviewer alike, that the study needs a revision of the figures.
To raise this wider awareness, we are about to publish a concise, but complete perspective on the importance of Scientific colour maps soon, and hope this will have some impact.
I hope this will have some impact. I already reviewed a paper where I complained about the used colormap. It never was changed because the editor just said it is ok this way.
I know; happened to me too. Extremely frustrating.
But all we as reviewers can do at the moment is keep pointing it out in the reviews and to our colleagues.