Color map
Not to confuse with color mapping in photography.
Color mapping (data visualization) is a fundamental practice used in data visualization to map numeric data values to unique colors (e.g., to create a type of visualization called heat map). The color map is a palette (or color scheme) of multiple individual colors and a fundamental property of color mapping. The variation in color along a color map may be by hue or intensity, giving obvious visual cues to the reader about the magnitude of the phenomenon. There are two fundamentally different ways of color mapping: the scientific color mapping and the unscientific color mapping.[1] Unscientific color mapping uses color maps, like the common rainbow color map, that have no uniform color gradient as seen by the human eye and are not accessible for color-vision deficient and color-blind readers. As such, color mapping is distorting the underlying data and not universally accessible. Scientific color mapping uses scientifically-derived color maps (or scientific color maps) with a uniform color gradient that can easily be ordered from one end to the other and contain colors that can be differentiated also by color-vision deficient readers. "Scientific color map" is a relatively new term, but its properties have been proposed as best practice over several decades.
In color mapping, a color gradient (i.e., a color axis) can replace a third position axis and conveniently display a three-dimensional function in form of a two-dimensional plot. The data variation across the 3rd dimension is only accurately represented if the distance (i.e., position variation) and the color variation across the data range covered by the color bar are constant.
History
Color mapping became more widely used in the 1990s, when graphics software started to have enough memory to display multiple colors in a single figure. The mathematically simplest color map involving multiple hues is "rainbow". "Jet", a common variant of "rainbow", is, for example, a simple piecewise linear function of row index with breaks at 1/8, 3/8, 5/8, and 7/8 of the color bar length. At one point, most common software programs therefore used a variant of rainbow as their default palette. Yet, "rainbow"'s rise in prominence was not without criticism. In a series of books on data visualization theory, the French cartographer Jacques Bertin (1918–2010) warned against using color hue to represent ordered data.[2][3] Indeed, for many decades, cartographers commented on such risks.[4][5][6][7][8] labelling rainbow-style spectral color schemes “illogical”.[9][10] In a more forceful way, Monmonier (2018) described the use of color hue to represent intensity as an indication that “the map maker either knows little about map design or cares little about the map user.” Although some instances were highlighted where "rainbow" could be useful in representing data,[11] the cartographers' consensus was clear. However, the impact across scientific disciplines has been slower.
From the Earth and space science community, Light and Bartlein (2004)[12] highlighted problems with red-green contrasts for color-blind readers and encouraged journals to improve color accessibility. In a plea to the Computer Graphics and Visualisation field, Borland and Taylor II (2007)[13] outlined the need to eliminate rainbow color maps in the article "Rainbow Color Map (Still) Considered Harmful". Thyng et al. (2016)[14] outlined guidelines for effective and accurate color map selection in oceanography, proposing a software package to be used. Recently, most efforts have been within Climate Science. Stauffer et al. (2015)[15] in a journal article, provided basic guidelines for using color more effectively in visualizations, building on earlier work.
An open letter addressing the continued use of the "rainbow" color scheme was issued within the climate community.[16] This work, along with related efforts by the author and colleagues, started the online "#endrainbow" campaign to discourage the use of rainbow-like color schemes. Despite the campaign's reach, its most notable impact has been limited to those in Climate and Earth Sciences (based on analysis of "#endrainbow" usage).
Finally, numerous recent online blog posts, like Crameri (2017)[17] and repeated commentaries[18][19] on the importance of more scientific and inclusive color use have effectively disseminated basic knowledge and awareness across various science communities (particularly online and within the data visualization field). However, despite this effort, it appears that this knowledge hasn't yet fully permeated all scientists and seems to be slowly reaching scientific leaders, publishers, and software designers.
Classes
Different classes of color maps are available to represent specific data sets more intuitively:[1]
- Sequential color maps are ideal for gradually varying, monotonic data with no significant or special values. A dataset of positive temperature values, for example, is best displayed with a sequential color map. Sequential color maps are the most versatile, can effectively represent most data sets, and are therefore the optimal default in visualization software, given their general characteristics.
- Diverging color maps focus on the center point of the color bar, or the extreme values on either side.[20] These maps are ideal for so-called bimodal datasets that have data points deviating from a central value, such as a dataset of positive and negative temperature values deviating from the zero value.
- Cyclic (or circular or orbital) color maps have no beginning or end. They are useful for displaying periodic data sets, such as angular data or data from Synthetic Aperture Radar (SAR) interferometry, which is used to display Earth's surface displacement. Cyclic color maps are diverging color maps with matching ends.
- Multi-sequential color maps consist of two or more sequential palettes arranged next to each other, with the direction of lightness increase being the same for all segments. An example is the "oleron" topography color map, which uses a bluish part for surface topography below the ocean and a greenish-brownish part for above the ocean, with the transition point fixed at sea level (i.e., the zero topography value).
Types
Different types of color maps help represent datasets more clearly:[1]
- Continuous color maps have a smooth color gradient. Designed for displaying sets of continuous, ordered data points, they can show both minor and significant data variations. A continuous color map usually uses over a hundred different color values. A continuous color map is generally the most versatile, and the other types (discrete and categorical) can be constructed from it.
- Discrete color maps feature a subset of color values from a continuous color map, that are clearly distinguishable from each other. While intended to visualize discrete data points, discrete color maps are also used with continuous datasets; they are frequently used with ordered data.
- Categorical color maps use multiple, unique, unordered colors to distinguish individual data points or entire graphs. They are applicable to various graph types including scatter plots, line plots, bar plots, and others.
Availability
While unscientific color maps are widely available in various color combinations, scientifically-derived color maps have only become more widely available in the last century. Pre-made scientific color maps are directly available to users, often integrated into common visualization toolboxes.
- ColorBrewer color maps developed by Cynthia Brewer.[21]
- MPL (Matplotlib) color maps developed by Stéfan van der Walt and Nathaniel Smith.[22]
- Cividis color map developed by Jamie R. Nuñez and colleagues.[23]
- CMOcean color maps developed by Kristen M. Thyng and colleagues.[14]
- CET color maps developed by Peter Kovesi.[24]
- Scientific color maps developed by Fabio Crameri.[25][26]
Reference list
- ↑ 1.0 1.1 1.2 1.3 Crameri, Fabio; Shephard, Grace E.; Heron, Philip J. (December 2020). "The misuse of colour in science communication". Nature Communications. 11 (1): 5444. Bibcode:2020NatCo..11.5444C. doi:10.1038/s41467-020-19160-7. ISSN 2041-1723. PMC 7595127 Check
|pmc=value (help). PMID 33116149 Check|pmid=value (help). - ↑ Hays, William L. (January 1985). "Review of Semiology of Graphics: Diagrams Networks Maps". Contemporary Psychology: A Journal of Reviews. 30 (1): 78. doi:10.1037/023518. ISSN 0010-7549.
- ↑ "Β. Graphic Constructions", Graphics and Graphic Information Processing, Berlin, Boston: DE GRUYTER, 1981, doi:10.1515/9783110854688.24, ISBN 978-3-11-085468-8, retrieved 2020-12-25
- ↑ "1. Introduction: Envisioning Semantic Information Spaces", Semantic Knowledge Representation for Information Retrieval, Berlin, Boston: DE GRUYTER, pp. 1–12, 2014, doi:10.1515/9783110329704.1, ISBN 978-3-11-032970-4, retrieved 2020-12-25
- ↑ MacDonald, Lindsay (October 1992). "Effective color displays — Theory and practice". Displays. 13 (4): 198. doi:10.1016/0141-9382(92)90032-m. ISSN 0141-9382.
- ↑ BREWER, CYNTHIA A. (1994), "Color Use Guidelines for Mapping and Visualization", Visualization in Modern Cartography, Modern Cartography Series, Elsevier, 2, pp. 123–147, doi:10.1016/b978-0-08-042415-6.50014-4, ISBN 978-0-08-042415-6, retrieved 2020-12-25
- ↑ Nelson, Elisabeth S. (1996-06-01). "How Maps Work: Representation, Visualization, and Design". Cartographic Perspectives (24): 27–30. doi:10.14714/cp24.757. ISSN 1048-9053.
- ↑ "Thematic Visualization", Web Cartography, CRC Press, pp. 165–194, 2013-12-10, doi:10.1201/b16229-13, ISBN 978-0-429-10754-2, retrieved 2020-12-25
- ↑ "Robinson, Arthur H. Elements of cartography. New York: John Wiley and Sons, Inc. 1953. 254 P. $7.00". Science Education. 38 (4): 317. October 1954. Bibcode:1954SciEd..38Q.317.. doi:10.1002/sce.3730380452. ISSN 0036-8326.
- ↑ Raveneau, Jean (1993). "Monmonier, Mark (1991) How to Lie with Maps. Chicago, University of Chicago Press, 176 p. (ISBN 0-226-53415-4)". Cahiers de géographie du Québec. 37 (101): 392. doi:10.7202/022356ar. ISSN 0007-9766.
- ↑ Brewer, Cynthia A. (January 1997). "Spectral Schemes: Controversial Color Use on Maps". Cartography and Geographic Information Systems. 24 (4): 203–220. doi:10.1559/152304097782439231. ISSN 1050-9844.
- ↑ Light, Adam; Bartlein, Patrick J. (2004). "The end of the rainbow? Color schemes for improved data graphics". Eos, Transactions American Geophysical Union. 85 (40): 385. Bibcode:2004EOSTr..85..385L. doi:10.1029/2004eo400002. ISSN 0096-3941.
- ↑ Borland, David; Taylor Ii, Russell M. (March 2007). "Rainbow Color Map (Still) Considered Harmful". IEEE Computer Graphics and Applications. 27 (2): 14–17. doi:10.1109/mcg.2007.323435. ISSN 0272-1716. PMID 17388198.
- ↑ 14.0 14.1 Thyng, Kristen; Greene, Chad; Hetland, Robert; Zimmerle, Heather; DiMarco, Steven (2016-09-01). "True Colors of Oceanography: Guidelines for Effective and Accurate Colormap Selection". Oceanography. 29 (3): 9–13. doi:10.5670/oceanog.2016.66. ISSN 1042-8275.
- ↑ Stauffer, Reto; Mayr, Georg J.; Dabernig, Markus; Zeileis, Achim (2015-02-01). "Somewhere Over the Rainbow: How to Make Effective Use of Colors in Meteorological Visualizations". Bulletin of the American Meteorological Society. 96 (2): 203–216. Bibcode:2015BAMS...96..203S. doi:10.1175/bams-d-13-00155.1. hdl:10419/101098. ISSN 0003-0007.
- ↑ Hawkins, Ed (March 2015). "Scrap rainbow colour scales". Nature. 519 (7543): 291. doi:10.1038/519291d. ISSN 0028-0836. PMID 25788088. Unknown parameter
|s2cid=ignored (help) - ↑ "The Rainbow Colour Map (repeatedly) considered harmful". Geodynamics. Retrieved 2020-12-25.
- ↑ Albrecht, Mario (2010-09-29). "Color blindness". Nature Methods. 7 (10): 775. doi:10.1038/nmeth1010-775a. ISSN 1548-7091. PMID 20885436.
- ↑ Wong, Bang (2011-12-28). "Erratum: Reply to "More on color blindness"". Nature Methods. 9 (1): 110. doi:10.1038/nmeth0112-110. ISSN 1548-7091. Unknown parameter
|s2cid=ignored (help) - ↑ Moreland, Kenneth (2009), "Diverging Color Maps for Scientific Visualization", Advances in Visual Computing, Lecture Notes in Computer Science, Berlin, Heidelberg: Springer Berlin Heidelberg, 5876, pp. 92–103, doi:10.1007/978-3-642-10520-3_9, ISBN 978-3-642-10519-7, OSTI 1141453, retrieved 2020-12-25
- ↑ "ColorBrewer: Color Advice for Maps". colorbrewer2.org. Retrieved 2020-12-25.
- ↑ "matplotlib colormaps". bids.github.io. Retrieved 2020-12-25.
- ↑ Nuñez, Jamie R.; Anderton, Christopher R.; Renslow, Ryan S. (2018-08-01). "Optimizing colormaps with consideration for color vision deficiency to enable accurate interpretation of scientific data". PLOS ONE. 13 (7): e0199239. arXiv:1712.01662. Bibcode:2018PLoSO..1399239N. doi:10.1371/journal.pone.0199239. ISSN 1932-6203. PMC 6070163. PMID 30067751.
- ↑ Kovesi, Peter (December 2015). "Bad Colour Maps Hide Big Features and Create False Anomalies". ASEG Extended Abstracts. 2015 (1): 1–4. doi:10.1071/aseg2015ab107. ISSN 2202-0586. Unknown parameter
|s2cid=ignored (help) - ↑ Crameri, Fabio (2020-01-06), Scientific colour maps, doi:10.5281/zenodo.1243862, retrieved 2020-12-25
- ↑ Crameri, Fabio (2018-06-29). "Geodynamic diagnostics, scientific visualisation and StagLab 3.0". Geoscientific Model Development. 11 (6): 2541–2562. Bibcode:2018GMD....11.2541C. doi:10.5194/gmd-11-2541-2018. ISSN 1991-9603.
New page for the fundamental visualization tool "color map".
Resubmit with adjustments to review comments
This article "Color map" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Color map. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
