Stat 6560 - Statistical Graphics and Visualization

TR 1:30-2:45
Engineering 206

Instructor:
Michael Minnotte
Office: Lund 201-C
Phone: 797-2844
E-mail: Mike.Minnotte@usu.edu
Office Hours: TR 9:30 - 10:20, W 10:30 - 11:20, or by appointment

Text:
Tufte, Edward R., (1983) The Visual Display of Quantitative Information, Cheshire, Connecticut, Graphics Press.
Cleveland, William S., (1993), Visualizing Data , Summit, New Jersey, Hobart Press.
S-plus scripts for figures
Data in S-plus format

Announcements

I will post announcements here. Please check regularly, especially if you must miss class for any reason.

Statistical graphics and data visualization are critical elements of modern data analysis and presentation. From initial exploration of a data set to the final presentation of results to the end user, graphics play a vital role in shaping our understanding of our data. Through proper use of graphics, we can make critical discoveries, and communicate them clearly. Conversely, poor use or misuse of graphics can seriously mislead (by accident or design).

In this course, we will start with presentation graphics, including discussion of both tools and principles which lead to clear communication and those which serve only to confuse or mislead. We will spend most of the semester in exploratory graphics and data analysis, including data mining. This will be broken down largely by the dimension of the applicable data. One- and two-dimensional datasets require and allow far different methods than those of more than three dimensions. Categorical and regression data call for their own specialized methods.

Even more than most aspects of statistics, graphics and visualization involve art as well as science. In most cases, there are many reasonable approaches. But an understanding of the options available and the underlying priciples will lead to successful analysis and presentation.

Prerequisites: I will not enforce any specific prerequisites. A course or two in traditional statistical analysis would be very helpful, as would prior computing experience, but neither is required. Previous experience with S-plus or R (statistical computing packages) will give a head start, but again is not necessary.

Other Sources: Beyond the required texts, material will be drawn from a number of sources. Some additional useful references are:
Cleveland, William S., (1994), The Elements of Graphing Data, Summit, NJ, Hobart Press.
Cleveland, William S. and McGill, Marylyn E., eds., (1988), Dynamic Graphics for Statistics, Belmont, CA, Wadsworth & Brooks/Cole.
Cook, R. Dennis and Weisberg, Sanford (1994), An Introduction to Regression Graphics, New York, Wiley.
Deboeck, Guido and Kohonen, Teuvo, eds., (1998), Visual Explorations in Finance with Self-Organizing Maps, New York, Springer.
du Toit, S.H.C., Steyn, A.G.W., and Stumpf, R.H., (1986), Graphical Exploratory Data Analysis, New York, Springer-Verlag.
Harris, Robert L., (1999), Information Graphics, New York, Oxford University Press.
Henry, Gary T., (1995), Graphing Data: Techniques for Display and Analysis, Thousand Oaks, CA, SAGE Publications.
Murrell, Paul, (2005), R Graphics, London, Chapman & Hall.
Playfair, William, (2005), The Commercial and Political Atlas and Statistical Breviary, New York, Cambridge University Press. (First published in 1786-1801!)
Robbins, Naomi B., (2005), Creating More Effective Graphs, Hoboken, NJ, Wiley.
Tufte, Edward R., (1990), Envisioning Information, Cheshire, CT, Graphics Press.
Tufte, Edward R., (1997), Visual Explanations: Images and Quantities, Evidence and Narritive, Cheshire, CT, Graphics Press.
Tukey, John W., (1977), Exploratory Data Analysis, Reading, MA, Addison-Wesley.
Unwin, A., Theus, M., and Hofmann, H., (2006), Graphics of Large Datasets: Visualizing a Million, New York, Springer. Wainer, Howard, (1997), Visual Revelations: Graphical Tales of Fate and Deception from Napoleon Bonaparte to Ross Perot, New York, Springer-Verlag.
Wallgren, A., Wallgren, B., Persson, R., Jorner, U., and Haaland, J., (1996), Graphing Statistics and Data: Creating Better Charts, Newbury Park, CA, SAGE Publications.
Westphal, Christopher and Blaxton, Teresa, (1998), Data Mining Solutions: Methods and Tools for Solving Real-World Problems, New York, Wiley.
Wilkinson, Leland, (1999) The Grammar of Graphics, New York, Springer.

Software

We will primarily be using R, a GNU-license statistical package and clone of S-plus. We will also be using GGobi for high-dimensional analysis. Both are available on the web for free download and can be used on Unix or Windows systems.

I will also provide a number of functions designed to be used in R. These will be posted here. You will need to download these functions, then pull them into R using the "source" command, after which you will be able to use these functions in your own analyses.

Graphical Statistics Software Sites
The Comprehensive R Archive Network
Windows R Setup Executable Download - click on rwXXXX.exe, where XXXX gives the version number
R Frequently Asked Questions (FAQ) List
R for Beginners (76 page pdf file)
An Introduction to R (100 page pdf file)
Data Analysis and Graphics Using R -- An Introduction (112 page pdf file)
Download site for GGobi
Download site for Mondrian

Note: If you are using R on campus, you will probably need to execute the following commands to get through the firewall to install additional packages:

Sys.putenv("http_proxy"="http://proxy.usu.edu:80/")
Sys.getenv("http_proxy")

You will probably need to run these commands (on the R console) at the beginning of your session before trying to install packages.

R Graphics Functions

Demonstration kernel density estimate R function
Mode tree R functions
Mode forest R functions
SiZer R functions
Multiple kde's broken down by a factor R function
Bivariate histogram function
RGL Perspective Plot R function
Local polynomial regression R functions
Trivariate kernel density estimation and plotting R functions
Trivariate local polynomial regression R functions
Andrews curve R function
Data image R function

Assignments

There will be a variety of assignments throughout the quarter. Each assignment will include a value (typically 20-100 points) that it will be scored out of. Your final grade will be determined by the sum of your points in all assignments. Some assignments will include combinations of analysis of existing graphics, creation of your own, computer work (mostly in S-plus, but some other packages/languages as well), and short oral presentations. The value of each assignment should be roughly proportional to its importance and the amount of work involved. Assignments will be handed out in class and posted here.

Assignment 1
Assignment 2
Assignment 3
Assignment 4 - Uninsured Data

Data

chondrite data
stamp thickness data
singer height data
melanoma rate data
ethanol data
Mount St. Helens earthquake data
high-dimensional data
PRIM7 particle physics data