Out of sight, out of mind
Download GGobi for Windows, Mac and Linux
GGobi is an open source visualization program for exploring high-dimensional data. It provides highly dynamic and interactive graphics such as tours, as well as familiar graphics such as the scatterplot, barchart and parallel coordinates plots. Plots are interactive and linked with brushing and identification.
GGobi is fully documented in the GGobi book: "Interactive and Dynamic Graphics for Data Analysis".
If you are interested in how GGobi came to be, you can read more about it on our history page.
- Need to look up cases with low or high values on some variables (price, weight,...) and show how they behave in terms of other variables? → brush in linked plots.
- Want to fluently examine the results of your R analyses in R? Use rggobi to easily transfer data between the two applications.
- Need to "see" the separation between clusters in high dimensions? → take a (helicopter) tour in high dimension.
- Want to examine network data? → lay out the network graphically and link it to other plots.
- Need to create high-quality publication graphics? Read about the describe display plugin and associated R package.
- Read up on more great features on the features page.
These methods have grown from exploratory data analysis, an area pioneered by John Tukey, building on the current fast graphically capable desktop computers to enable highly interactive, real-time, dynamic exploration of data.
Who is using GGobi?
- High dimensional geometric shapes, by Barret Schloerke.
- ASA Data Expo 2006: "Glaciers Melt as Mountains Warm" with Hadley Wickham, Jonathan Hobbs, Dianne Cook
- Info Vis Contest 2005:"Boom and Bust of Technology Companies at the Turn of the 21st Century"
with Hadley Wickham, Dianne Cook, Junjie Sun, Christian Röttger
- The RGtk2-based rattle by Graham Williams uses rggobi for some graphics
- exploRase, a tool for exploratory multivariate analysis of systems biology data, uses rggobi and ggobi for data display
- Classifly, an R package that uses rggobi to explore classification boundaries in high-dimensional spaces
- SpikeOMatic, an R package which also uses GGobi to do spike sorting. Christophe Pouzat.
- Visualization of comparative genomic analyses by BLAST score ratio. David A Rasko, Garry SA Myers and Jacques Ravel.
- From paragraph to graph: Latent semantic analysis for information visualization. Thomas K. Landauer, Darrell Laham and Marcia Derr.
- Clustering by Local Skewering (PDF). David W. Scott.
GGobi might not be exactly what you need, so have a look at Mondrian and Manet both produced by the RoSuDa group at the University of Augsburg. Currently, their categorical features are much stronger than GGobi's so be sure to check them out if you are analysing heavily categorical data.