History of Charts: Who Invented This Chart Type?

Joseph Priestley (1733 – 1804) is credited with the first timeline charts using individual bars – to visualize the life span of a person – in 1765. Priestley’s timelines proved a commercial success and a popular sensation, and went through dozens of editions.

Twenty years later, William Playfair invented the bar, line and area charts, using his work as main source of inspiration.

William Playfair (1759 – 1823) is credited with the inventions of the bar, line and area charts in 1786. His time-series plots are still presented as models of clarity.

He was a Scottish engineer and political economist. Playfair is considered the founder of graphical methods of statistics. Playfair argued that charts communicated better than tables of data.

William Playfair (1759 – 1823) – the creator of bar, line and area charts – invented also the pie chart and circle graph, in 1801.

First pie chart was published by Playfair in his Statistical Breviary, in London. Ir was showing the proportions of the Turkish Empire located in Asia, Europe and Africa before 1789. Here are other examples:

André-Michel Guerry (1802 – 1866) was the first to publish a map (of France) with regions using different colors, based on some variable (crime levels). In 1829 he created the first Choropleth Map.

Together with Adolphe Quetelet, André-Michel Guerry – a French lawyer and amateur statistician – may be regarded as the founder of moral statistics.

Florence Nightingale (1820 – 1910) – an English social reformer and statistician, the founder of modern nursing – is credited with developing a form of pie chart she first called coxcomb, after 1856.

Now it is known as Polar Area diagram, or even Nightingale Rose diagram. Polar area charts have been also used before, by André-Michel Guerry in 1829, and Léon Lalanne in 1843.

Georg von Mayr (1841-1925), a German scientist, used first polar diagrams and star plots (i.e. radar or spider web plors) for data representation, in 1877. Georg von Mayr is also known for the first use of proportional, divided square in the modern mosaic form for data representation.

Other people – like André-Michel Guerry, Léon Lalanne or Florence Nightingale – already used before a polar plot.

Philbert Maurice d’Ocagne (1862 – 1938), in 1885, is known to have described a method with “coordonnées parallèles” in his book. However, the method the way we know it today has been rather popularized years later by Alfred Inselberg, in 1959.

Henry Gannetts in 1880, and Henry Gannetts in 1898, are also known for using and describing some method with parallel coordinates.

Karl Pearson (1857 – 1936), an English mathematician and bio-statistician, credited with establishing the mathematical statistics, is also known as having introduced the term or “histogram” in 1891.

The term is believed to be coined from “historical diagram”. Even if sometimes it is said to be derived from Ancient Greek.

H. Riall Sankey (1853 – 1926), an Irish engineer and captain in the Royal Engineers, was the first known to have used this type of chart in 1898, in a classic figure showing the energy efficiency of a steam engine. Sankey diagrams have been later named after him.

But Charles Minard’s “Map of Napoleon’s Russian Campaign of 1812” – a flow map overlaying a Sankey diagram onto a geographical map – was created in 1869 and is one of the most famous Sankey diagrams.

Henry Gantt (1861 – 1919), an American mechanical engineer and management consultant, created this type of chart around 1910.

Gantt charts were employed on major infrastructure projects and continue to be an important tool in project and program management.

Thomas L. Saaty (1926 – 2017), former professor at the University of Pittsburgh, used in 1964 arc diagrams (or linear embeddings) to study crossing numbers of graphs. T.A.J. Nicholson used them too, in 1968.

A similar type of diagram was used by M. Wattenberg in 2002, to visualize the repetition patterns in strings. He used arcs to connect pairs of equal substrings.

John Tukey (1915 – 2000) was an American mathematician who introduced the box plot in his “Exploratory Data Analysis” 1977 book . He is also credited with coining the term “bit”.

Tukey’s box plot assumes symmetry for the whiskers and normality for their length.

William S. Cleveland, computer scientist and professor, invented the dot plots back in 1981, as a replacement for bar graphs or pie charts, which were using too much real estate and clutter.

Not to be confused with the pre-computer era Dot Plot charts from statistics, which has been around since 1884, in which each bar is replaced by a series of dots.

Edward Tufte is credited with being the first who formally documented this type of small embedded chart, in 1983, in his book “The Visual Display of Quantitative Information“. He named it “sparkline” in 2006. It’s fair to say he invented only the name and popularized it as technique.

The earliest known implementation is of Peter Zelchenko, in 1998. In 2008, Microsoft filed a patent application for the implementation of sparklines in Microsoft Excel 2010.

Edward Tufte, also in his 1983 book “The Visual Display of Quantitative Information“, described the Slope Chart or the Slope Graph as a new type of chart. The term “slopegraph” was actually coined around 2011.

Slope charts display exactly two points in time and visualize if the variations between them went up or down, using simple straight lines.

Ben Shneiderman, from the University or Maryland, looked for a better way to visualize the files and folders from his disk drive. In 1990, he tried to do it with a tree diagram, but there were too many nodes and branches.

The treemap was his solution, with less clutter and fully nested rectangles, each area proportional with the represented amount.

Jonathan Feinberg created the first word-cloud layout algorithms back in 2005, while working on a social bookmarking application at IBM. In 2008, he created the web application Wordle, the first and best-known app from this family.

The app is still online today, but looks a bit obsolete, as it is using Java applets. There are however dozens of other online word clouds today, or word cloud chart algorithms, available.

Stephen Few developed the bullet graph in 2006, in his book “Information Dashboard Design: The Effective Visual Communication of Data“, to replace the rich meters and gauges that are often used on dashboards.

Stephen was very upset with the amount of clutter traditional dashboard had at that time. His simple linear bullet graph renders bars, which are easier to read and estimate by the human eye than circular angles or arc lengths.

Martin Krzywinski, a data scientist, published an article with the first known Chord Diagram in the New York Times, in 2007. It was created with Circos, a well-known application today for this type of visualizations.

The article was called “Close-ups of the Genome, Spieces by Spieces by Spieces” and it is still available online.

Lee Byron developed and published the first Stream Graph in a New York Times article, in 2008. It represented data about movie box office revenues in a new way.

The Stream Graph is in fact a variation of a Stacked Area Graph, with values displaced around a varying central baseline. It resembles a river-like stream and this makes it aesthetically pleasing and more engaging to look at.