GGobi

from Wikipedia, the free encyclopedia

GGobi is a free statistics software to visualize and analyze high-dimensional multivariate data with dynamic graphics. GGobi can be fully integrated into the R programming language, which is also free . Brushing and linking of the data points is supported within the GGobi graphics.

overview

GGobi was developed to visualize high-dimensional data matrices. The developers were interested in researching high-dimensional data structures. The name of the software changed several times over the years, especially because the underlying technology changed:

  • Mid 1980s: Dataviewer,
  • 1989: XGobi and
  • from 1999: GGobi.

GGobi is published under a license that is a combination of three free licenses , so GGobi is free software .

2D projection ( Grand Tour ) of a six-dimensional data set. Three clusters are visible and two data points are highlighted in yellow.
Parallel coordinate plot linked to the Grand Tour plot. The two data points marked yellow are also marked yellow here.

Graphics

Graphics can reveal more information about a distribution than metrics. Various tools in GGobi support the discovery of clusters, nonlinear structures, outliers and other structures in high-dimensional data.

Variables) Graphic (s)
1 Average Shifted histogram , dot plot , bar chart , Spineplot
2 Scatter plot
> 2 Scatterplot Matrix , Parallel Coordinates , Grand Tour , Projection Pursuit (guided and manual)

Interactivity

With the help of the interactive tools, data points or clusters can be marked:

  • Brushing
When the brush rectangle goes over a data point, it is displayed with a previously selected symbol and / or color. If the persistent option is selected for the brush, the data point retains its color and symbol even if it is no longer in the brush rectangle. Otherwise, its icon and color will be displayed as originally.
  • Identify
When the cursor comes over a data point, a text or variable value is displayed.
  • Linking
If the data points of a data set are displayed in different graphics at the same time, then changing a symbol or a color of a data point in a graphic leads to corresponding changes in all graphics.

literature

  • A. Buja, D. Cook, DF Swayne: XGobi: Interactive Dynamic Data Visualization in the X Window System. In: Journal of Computational and Graphical Statistics. Volume 7, No. 1, March 1998, pp. 113-130.
  • A. Buja, DT Lang, DF Swayne: GGobi: Evolving From XGobi into an Extensible Framework for Interactive Data Visualization. In: Journal of Computational Statistics and Data Analysis. Volume 43, No. 4, August 2003, pp. 423-444.
  • D. Cook, DF Swayne: Interactive and Dynamic Graphics for Data Analysis . Springer, 2007, ISBN 978-0-387-71761-6 .

Web links

Individual evidence

  1. ↑ ggobi Licenses page . Retrieved December 31, 2010.