Statistical Computing, Stat 244

MWF 9-10 330 Evans

Instructor: Phil Spector, Evans 495

Office Hours: MWF 10-12, T 2-5

email: spector@stat.berkeley.edu (spector on SCF systems)

This course is designed as a ``Special Topics'' course, covering several different applications of computing within the field of statistics. In addition to writing programs in a language such as C, C++ or Java, the course will emphasize use of existing subroutine packages and computing environments to understand the nature of statistical computing problems. Topics to be covered include:

• Multivariate Analysis
• Density Estimation and Smoothers
• Time Series
• Linear Models

In addition, a number of special topics in programming will be discussed, such as:

• Structured Programming
• Object Oriented Programming
• Statistical Graphics
• Recursion
• Statistical Package Design
• Text processing

Additional topics will be covered based on availability of time and interests of the class.

Students should have some knowledge of UNIX and a high level programming language, preferably C, C++ or Java. Additionally, students should have an understanding of the fundamentals of matrix manipulations and basic statistical methods.

There will be four computer projects assigned at regular intervals throughout the course based on the following four topics: multivariate analysis, density estimation/ smoothing, time series and analysis of variance. If you have a special project of interest which is related to topics covered in class, feel free to suggest it as an alternative to an assignment given in class.

All students will be provided with a computer account on the Statistical Computer Facilities (SCF) network of SUN Microsystems computers. The computer room in 491 Evans provides high resolution workstations; you can also remotely log in to the SCF system from other campus computers using the SSH protocol. If you wish to do your assignments on some other computer, keep in mind that required programs may be stored on the SCF system, and it is your responsibility to get the programs to another computer. Additionally, some of the assignments require the use of UNIX utilities or programs, so if you wish to use a non-UNIX computer, you should make sure that suitable alternatives are available.

None of the following texts are required, but interested students may want to consider the following books, not only for this course, but as a useful part of their professional libraries:

1. Gentle, James E.: Numerical Linear Algebra for Applications in Statistics, Springer, New York(1998)
2. Gentle, James E.: Random Number Generation and Monte Carlo Methods, Springer, New York(1998)
3. Kernighan, Brian W. & Ritchie, Dennis M.: The C Programming Language, Prentice Hall, New Jersey(1988)
4. Kernighan, Brian W. & Pike, Rob.: The UNIX Programming Environment, Prentice Hall, New Jersey(1984)
5. Kennedy, William J. & Gentle, James E.: Statistical Computing, Marcel Dekker, New York (1980)
6. Thisted, Ronald A.: Elements of Statistical Computing, Chapman and Hall,New York (1988)
7. Press, Williiam H. et. al.: Numerical Recipes in C, Cambridge University Press, Cambridge (1988)