.png)
Astreamgraph, orstream graph, is a type of stackedarea graph which is displaced around acentral axis, resulting in a flowing, organic shape. Unlike a traditional stacked area graph in which the layers are stacked on top of an axis, in a streamgraph the layers are positioned to minimize their "wiggle". More formally, the layers are displaced to minimize the sum of the squared slopes of each layer, weighted by the area of the layer.[1] Streamgraphs display data with only positive values, and are not able to represent both negative and positive values.[2]
Streamgraphs and their use were popularized byAmanda Cox in a February 2008New York Times article on moviebox office revenues.[3] Cox got the idea from then-undergraduate Lee Byron,[1] who had used a similar method for visualizing his music listening history.[4]
A related graph, sometimes conflated with streamgraphs, is the ThemeRiver, in which the "silhouette" of the graph is symmetrically arranged around the central axis.[1]
Streamgraphs were found to be more readable than basic stacked area graphs or ThemeRivers for value comparison tasks.[5]
Streamgraphs are officially supported byMatplotlib[6] andD3.js.[7]
Marco Di Bartolomeo and Yifan Hu (2016) propose several improvements to streamgraphs, such as using1-norm minimization instead of2-norm minimization.[2]
 | Thisstatistics-related article is astub. You can help Wikipedia byexpanding it. |