ASA Statistical Graphics Video Library

The graphics video library is an archive of great technical and historic interest which captures much of the history of dynamic graphics for data analysis over the past 30 years. The collection is currently composed of 38 videos, listed below in chronological order.

Multidimensional Scaling

J.B. Kruskal, AT&T Bell Laboratories (1962)

Image of a Thunderstorm
Anne Freeny and John Gabbe. AT&T Bell Laboratories (1966)
This sample frame (which is, fortunately, much clearer in the video) shows the intensity of a storm in different quadrants of a geographic area. Primarily of historical interest, although the narrator makes the refreshing admission that the technique used in the video was not found to be especially useful.

Real-time Rotation
Jih-Jie Chang. AT&T Bell Laboratories (1970)

J.W. Tukey, J.H. Friedman and M.A. Fisherkeller. Stanford Linear Accelerator (1973)
prim9.gifIn this sample frame from the video, John Tukey sits in front of the Prim-9 hardware -- and the blackboard he uses in his explanation of the variables in the particle physics data.
prim9a.gifThere's no good excuse for including this frame, but it's irresistible somehow. This video could be of great interest as part of a 'statistics day' program, and could also be useful for a statistical computing or advanced graphics class.

Ozone in the Northeast
Richard A. Becker, William S. Cleveland, Beat Kleiner & Jack L. Warner,AT&T Bell Laboratories. (1978)

The video shows the regional nature of ozone transport in the Northeast US -- you can see ozone concentrations carried from the New York area into central Connecticut and Massachusetts, resulting in high ozone concentrations there, even though these areas do not produce the ozone themselves.
Ozone concentrations are represented by circle size (area?). The circle making up the clock is the size of the EPA standard for ozone.
The movie was shot frame-by-frame from a monitor on a DDP 224 computer running the GR-Z graphics software (the graphics subroutine library underlying S). Jack Warner did the coding and data management; Rick Becker was involved with the GR-Z library; Bill Cleveland and Beat Kleiner were responsible for the data analysis.

Brushing a Scatter Plot Matrix

brushing85a.gifThe same equipment and people were involved as in Dynamic Displays of Data ; this video was shot and produced at that same time. It contains several data examples (Fisher's Iris Data, Rubber Abrasion Loss, Brain/Body weight) illustrating the capabilities of the brushing software.
This could be usefully shown to a class of novices at interactive statistical graphics. (Note: The content overlaps with that of Dynamic Displays of Data. )

Dynamic Displays of Data
Richard A. Becker and Robert McGill. AT&T Bell Laboratories (1985)
dyndisp85a.gifHere sits Rick Becker pointing at the screen of the AT&T terminal that was used in this video.

dyndisp85b.gifThis video tape is a tutorial that shows various kinds of dynamic displays. It was made with a professional video crew filming people and displays on 3/4" video tape. Post-production was done on 1" reel-to-real video tape. Bob McGill was director of the video production; Rick Becker did the on-screen narration and the software. The equipment was an AT&T Teletype 5620 bit-mapped display terminal that ran a Motorola 68000 CPU internally. The 5620 ran all of the dynamic graphics programs; they were written in C, cross-compiled, and downloaded into the 5620 for execution. Below the terminal was an AT&T 3B2 computer running Unix and S -- the equipment shown on the desk was the total needed to support the application.

The tutorial shows dynamic point identification, dynamic power transformations, regression leverage in which a point is moved and the least-squares line is recomputed in real time, a mouse-controlled straightedge, 3-d point cloud rotation, and brushing a scatterplot matrix. There is a lot of motivational material, including 3-D models. It also has an interesting soundtrack of computer-generated music. This video was made to show dynamic graphics techniques that WERE NOT primarily 3-D rotation, which had been the main focus of dynamic statistical graphics from the time of Prim-9.

 Data Analysis Networks in DINDE
R.W. Oldford and S.C. Peters. University of Waterloo (1986)

In this video, Wayne Oldford describes the functionality of DINDE, an object-oriented graphical system for conducting and monitoring a data analysis session. This video is primarily of interest for those who are studying the design of statistical software systems.

Plot3d : Data Animation on a Sun Workstation
Paul Tukey and Vonn Marsch. Bell Communication Research (1987)

plot3d1.gifThis video is a demonstration of point cloud rotation, using data collected at Bellcore. The inset at the lower left of the screen shows the keyboard interactions; a nice touch.
plot3d2.gifIn this case, the inset shows the use of the mouse to control the rotation. The narration in this video discusses the data in some detail, so it could be an effective video for a classroom presentation about rotation.

Use of the Grand Tour in Remote Sensing
John A. McDonald and Steve Willis. University of Washington (1987)

Here's the initial image showing the confluence of two rivers.

Here's the mouse is used to select pixels from of one of the regions of interest: the transept of the river where waters of the two tributaries have not yet totally mixed.

Now the multivariate array is displayed using the grand tour in Dataviewer.
This video uses Dataviewer, software written by Andreas Buja and Catherine Hurley, to analyze Landsat data. It starts with an image of the confluence of two rivers near Manaus, Brazil. The area shown includes forest and grassland. Pixels are selected from several areas of the image, and these form an array of data: each pixel is represented by a longitude, a latitude, and four spectral bands. The grand tour is used to view the resulting data.
This video is primarily of interest for those who are working with image data, or learning about applications of high-dimensional rotation methods.

Odds Plots: Finding Associations between Views of a Data Set
Werner Stuetzle. University of Washington (1987)

This video uses Stuetzle's Plot Windows system on a Symbolics LISP workstation. The video contains an early presentation of linked plots, using histograms as well as scatterplots, and there is a careful description of the use of brushing to find high-dimensional relationships in the data. (The video uses the "Swiss fertility data," data on breast cancer survival, and data on the Mount St. Helens earthquake.)


These frames show the use of an odds plot to highlight structure which is very difficult to discern in scatterplots, even with the advantages of linked views. This video is of particular interest to people who would like to explore ways of representing or using odds plots.

Antelope: data Analysis with Object Oriented Programming and Constraints
John McDonald. University of Washington (1987)

The story of the data analysis is nicely told, and there are some unusual methods in use. McDonald has linked a multiway contingency table to a scatterplot of latitude and longitude so that various regional vowel pronunciations in Britain can be explored. He rearranges the rows and columns in a two-way table both automatically and interactively, and describes the method as an analog of rotation.

Automatic Tracing of Ice Floes on Satellite Images
Jeff Banfield. University of Washington (1987)

Dataviewer: A Program for Looking at Data in Several Dimensions
Andreas Buja and Paul Tukey. Bell Communication Research (1987)

This video contains a discussion of Dataviewer, interactive graphics software running on a Lisp workstation.

The window at the center of this second figure contains a rescaling of the data in the the figure above. Different observations about the data can be made from each figure.

Brushing and Rotation on an Iris

Richard A. Becker, William S. Cleveland and Gerald Weil
AT&T Bell Laboratories (1987)

This video describes and demonstrates some interactive graphics functionality on a Silicon Graphics Iris workstation, with an emphasis on showing the power of the workstation itself.

The narration is not as strongly focussed on the data as the earlier AT&T Bell Labs videos, so it is probably less suitable for classroom discussion.

Graphical Programming
G. D. DesVignes and R. W. Oldford University of Waterloo (1988)

This video describes the use of graphical programming with an example, showing the encapsulation of several steps of an analysis into a single reusable tool. A LISP environment is used for software development, probably a Xerox LISP workstation.

Higher Hierarchical Views of Statistical Objects
C. Hurley and R.W. Oldford. University of Waterloo (1988)

This video presents a very general discussion and demonstration of linked views: for instance, boxplots and scatterplots are linked. A LISP environment is used for software development, probably a Xerox LISP workstation. This video is primarily of interest for those who are studying the design of statistical graphics software.


Marilyn Becker, Linda A. Clark, and Daryl Pregibon. AT&T Bell Laboratories (1989)

This video is a tutorial for the use of a set of S functions for growing and displaying trees, as well as interacting with them in various ways.

Visualizing Multivariate Structure with VISUAL/Pxpl

Forrest W. Young & Penny Rheingans. University of North Carolina at Chapel Hill (1990)

This video describes the search for meaningful clusters in 6-dimensional space using data on crime rates in each state of the United States. The system, the methods and the data are described in Dynamic graphics for exploring multivariate data, by Forrest W. Young, Douglas P. Kent, and Warren F. Kuhfeld in Dynamic Graphics for Statistics, by W. S. Cleveland and M. E. McGill.

Thin Plate Splines and the Analysis of Biological Shapes

Fred L. Bookstein and William Jaynes. University of Michigan Medical Center (1990)

This video describes the use of splines on "landmarks" in images of brains to make it possible to make comparisons between brains of a variety of shapes and sizes.

This video is very polished and would be excellent both for the classroom and for 'statistics day' presentations.

XGobi: Dynamic Graphics for Data Analysis
Deborah F. Swayne, Dianne Cook, and Andreas Buja. Bellcore (1991)

XGobi is interactive software for exploratory data analysis. This video is an introduction to XGobi which briefly presents its main functions -- linked brushing, the labelling of points, interactive scaling, rotation, grand tour.

Focusing & Linking as Paradigms for the Visualization of High-Dimension Data
Andreas Buja, Bellcore, and John McDonald, John Michalak, Werner Stuetzle and Steve Willis. University of Washington (1991)

In this video the authors discuss two principles of interactive data visualization: 1) Focusing and 2) Linking views of data. Focusing refers to the methods for controlling individual views. Linking refers to methods for relating multiple views to each other. Examples of focusing include controls for zooming, panning, projecting.

A Simple Dynamic Graphical Diagnostic Method for Almost Any Model 

George S. Easton. University of Chicago (1991)

Visualizing Panel Data
Martin Koschat and Deborah F. Swayne. Bellcore (1992)

This video shows the use of XGobi to explore panel data. In addition to scatterplots, a plot called the "case profile" is used -- it's a parallel coordinates plot of a small number of cases. The data explored represent hourly telephone usage (number of calls placed) over some weeks or months, and the case profile is used to display time series plots.

Smoother's Workbench
Lise Manchester. Dalhousie University (1992)

SeeNet: See a Network

Richard A Becker, Allan Wilks, and Steven Eick. AT&T Bell Labs (1992)

Tokyo Data Map

Planned by the Tokyo Government. Produced by Dentsu Inc. and Dentsu Prox Inc. (1992)

Grand Tour and Projection Pursuit
Dianne Cook, Andreas Buja, Javier Cabrera, and Deborah F. Swayne. Bellcore (1993)

Manufacturing Process Data
William F. Eddy and Audris Mockus. Carnegie-Mellon University (1993)

Edge Information at Landmarks in Medical Images
Fred L. Bookstein and William D. K. Green. University of Michigan Medical Center (1993)

Incidence of Disease Mumps
William F. Eddy and Audris Mockus. Carnegie-Mellon University (1994)

Display of U. S. Air Traffic 

William F. Eddy and Shingo Oue. Carnegie-Mellon University (1994)

Exploring Time Series Using Interactive Graphics

Andrew McDougall and Dianne Cook. Rutgers University (1994)

Integrated Documents

Günther Sawitzki; Universität Heidelberg (1994,1999)

Computer Graphics in Statistics: The Lats 30 Years in Brief
Dianne Cook. Iowa State University (1995)

Spatial CDF Estimation & Visualization with Application to Forest Health Monitoring
James J Majure, Dianne Cook, Noel Cressie, Mark Kaiser, Soumendra Lahiri, Juergen Symanzik. Iowa State University (1995)

An Interactive Environmental for the Graphical Analysis of Spatial Data

James J. Majure, Dianne Cook, Juergen Symanzik, and Inna Megretskaia. Iowa State University (1995)

Dynamic Graphics in a GIS: Analyzing and Exploring Multivariate Data

Juergen Symanzik, James J Majure, Dianne Cook. Iowa State University (1995)

Missing Data in Interactive High-Dimensional Visualization
Deborah F Swayne, Bellcore; Andreas Buja, AT&T Bell Labs. (1996)

Dynamic Statistical Graphics in a Highly Immersive Environment

Juergen Symanzik, Dianne Cook, Bradley D. Kohlmeyer, Uli Lechner, Carolina Cruz-Neira. Iowa State University (1996)

A Tribute to J. Bertin's Graphical Data Analysis
Günther Sawitzki; Universität Heidelberg (1997,2014)

Among the rich material on graphical presentation of information in "La Graphique et le Traitement Graphique de l'Information" (1977), engl. "Graphics and Graphic Information Processing" (1981), Jaques Bertin discusses the presentation of data matrices, with a particular view to seriation. "A Tribute to J. Bertin's Graphical Data Analysis", presented at the SoftStat Conference '97, gives an appraisal of this aspect of Bertin's work. Bertin's approach has been implemented in the Voyager system for data analysis. This is a video of the SoftStat '97 presentation using Voyager. Video recorded in Feb. 2014.

Exploratory Data Analysis with Mondrian-The Tour de France
Antony Unwin, Universität Augsburg (2014)

LDAvis: A method for visualizing and interpreting topic models
Carson Sievert, Iowa State University (2014)

This video (recorded September 2014) shows how interactive visualization is used to help interpret a topic model using LDAvis. LDAvis is an R package which extracts information from a topic model and creates a web-based visualization where users can interactively explore the model. More details, examples, and instructions for using LDAvis can be found here.

Video Library Bibtex Citation

  author="{ASA Statistical Graphics Section}",
  title="Video {L}ibrary",