AMSC 660 / CMSC 660 Scientific Computing I, Fall 2006


Dianne P. O'Leary

oleary@cs.umd.edu

When and Where: TuTh......9:30am-10:45am (CSI 3120) (CSI is the Computer Science Classroom building, attached to A.V. Williams and behind the Wind Tunnel.)

Office Hours: Tuesday 8:45-9:15, Thursday 8:00-9:15, Friday 9-10, and by appointment, in AVW 3271. Email is welcome anytime. My other course meets TTh 11:00-12:15.

Teaching Assistant: Geping Liu (geping@math.umd.edu)
Office hours Tuesday and Thursday 3:00-5:00, AVW 3457.

Textbook: I'll make a draft of a textbook that I am writing for this course available to you in pdf files. Click here.

Prerequisite: Undergraduate numerical analysis. Programming assignments will be in Matlab.

Topics: Monte Carlo simulation, numerical linear algebra, nonlinear systems and continuation method, optimization, ordinary differential equations. Fundamental techniques in scientific computation with an introduction to the theory and software for each topic.

Grading: Based on quizzes, homeworks, and project.

Final Exam: None.

CMSC Masters Comprehensive Exam grades: based on best 7 of 9 quizzes.

Scientific Computing Certificate Program: If you are not an AMSC or CMSC major, then you may obtain a Certificate in Scientific Computing notation on your transcript by completing this course plus 661 and 662. Further information.

Basic Information:

  • Course Information and Syllabus
  • The textbook: Scientific Computing: A Second Course, with Case Studies
  • Title Page
  • Table of Contents
  • Preface
  • Bibliography for the textbook (Caution: the reference numbers might change later in the semester, and please ignore typos in the current version.)
  • Relevant units from the textbook Parts 1-6 are now posted.
  • Solutions for challenges and exercises in the textbook
  • UMCP Code of Academic Integrity
  • Information about computer accounts. You should now be able to access your GRACE account. For your assignments, you may use GRACE or any other machine with Matlab access.
  • Survival Guide for Scientific Computing
  • Grades after 9 quizzes and 3 homeworks

    • Lecture Notes:

  • Errors and Arithmetic (postscript)
  • Dense Matrix Computations
  • Monte Carlo Computations The Traveling Salesperson slides that I showed in class can be found in the 2004 notes.
  • demorand.m demonstration
  • Optimization
  • (Optional) In response to requests for more information about conjugate gradients, you can try either my notes or notes by J. R. Shwechuk.
  • Nonlinear Equations
  • Nonlinear Equations Case Study
  • ODE notes from AMSC/CMSC 460
  • ODE notes

    • Quizzes:

  • Answers to Quiz 1: Thursday, September 7.
  • Answers to Quiz 2: Tuesday, September 19. Note that many of you found this quiz difficult. Please stop by during office hours if you need help on this or anything else.
  • Answers to Quiz 3: Thursday, September 28.
    The figure was created by q3.m Please stop by during office hours if you need help on this or anything else.
  • Answers to Quiz 4: Tuesday, October 10. Known error on p182 and p184: The formula for the variance of the estimates of the integral for Monte Carlo are only correct for regions whose volume is 1. (The same error occurs in the lecture notes.)
  • Answers to Quiz 5: Thursday, October 19.
  • Answers to Quiz 6: Tuesday, October 31.
  • Answers to Quiz 7: Thursday, November 9. (On Nonlinear equations notes pp 1-10 and Chapter 23 through p. 275.)
  • Answers to Quiz 8: Tuesday November 21. (On first part of 460-ode material. Some of it is covered in the textbook.)
  • Answers to Quiz 9: Tuesday December 5.

    • Homework:

  • Homework 1: due September 19, 1pm.
    (Late penalty applies at 1:01pm on September 19.)
    Complete the 5 challenges in Chapter 2 of the text. Show all work. Hand in your written work, listings of your Matlab programs, "diary" listings of the output, and copies of the plots. Document your Matlab programs as discussed on p.36 of the text and as illustrated in the sample programs. Answers to frequently asked questions:
  • In each random trial for Challenge 2.4b, the birth/death and migration rates change each year.
  • The Lagrange multiplier for each constraint is the derivative of the function value with respect to changes in the constraint, as discussed on p.20.
  • In Challenge 2.5, assume that the error components are independent, so that S = tau I.
  • The "first unit vector" in 2.5 is the first column of the identity matrix, defined in the pointer on p.48.
    • Points:
  • Challenge 2.1: 8 points
  • Challenge 2.2: 8 points
  • Challenge 2.3: 8 points
  • Challenge 2.4: 10 points
  • Challenge 2.5: 6 points
  • Look for the solution here.

  • Homework 2: due October 17, 1pm.
  • The homework
  • myf.m
  • FAQ Check these for point values and (important!) which challenges to skip, as well as corrections to the homework.
  • Solution

  • Homework 3: due November 28, 1pm.
  • The homework
  • beetledata.m The data.
  • FAQ
  • Solution

  • Homework 4: due December 12, 1pm.
  • The homework
  • basiccode.m
  • FAQ
  • New: Solution


    • Term Project Information

    Some Information from Fall 2004