Physics 317: Introduction to Computational Physics and Programming

Course material can be found online at URL

Instructor

Michael Richmond
Building 76, Office 1274
Office phone: 475-2538
E-mail: mwrsps@rit.edu

Syllabus

Class meeting times

        T            4:00 -  5:50  PM       08-1365   (Gosnell Mac Lab)
        R            4:00 -  5:50  PM       08-1365   (Gosnell Mac Lab)

Students must appear at the Tuesday session, during which I will introduce the topics of the week, describe the week's project, and explain techniques.

Students must attend the Thursday lab sessions unless they have finished the week's assignment already. I will count attendance in determining the final grades.

In order to submit your work each week, you must do the following:

1. Source code file(s) must be submitted via the myCourses WWW page. After you have logged into myCourses, go to the Phys317 page and then choose the Dropbox item. Place your code in the appropriate folder -- there should be one per assignment.

2. All other material -- answers to specific questions, tables, figures, analysis, etc. -- must be printed or written on paper. You can hand the papers to me, or place them into the mailbox outside my office door. If you submit multiple pieces of paper, please use staples or paper clips to fix them together.

Each week's assignment will be graded on a scale of 0 to 10. I will assign points roughly in the following manner (though there may be small variations from week to week):

• pseudocode submitted before class on Wednesday: 1 point
• assignment submitted on time: 2 points
• code runs without error messages, given proper input: 2 points
• code computes the desired quantities accurately: 3 points
• description/explanation of work: 2 point
• HTML man page explaining the usage and purpose of the code: 1 point

All programs should include

• lines no longer than 80 characters, even after tab characters are expanded into spaces
• comments at the start of each function, explaining the purpose of the function. Each function argument should have a separate comment, describing its meaning, its type, valid ranges, and whether it is an input or output to the function.
• comments sprinkled throughout the source code, roughly at the start of each "paragraph"

Lectures

Week 1
• Examples of Numerical Attacks on Hard Problems
• Introduction to Scilab
• Your first Scilab programs
• Functions and Plotting in Scilab
• Programming Style
• Pseudocode: What it is, and why you should use it

  Week 2

• How computers represent of numbers
  • Answers to Challenges
• Error, accuracy, and precision
• Dangers of computer arithmetic
• An example of accumulated roundoff error

  Week 3

• Open Methods for finding roots
• Closed Methods for finding roots
• Using Functions as arguments

  Week 4

• Blackbody radiation and the Planck Function
• Numerical Integration: Basic
• One Debugging Technique

  Week 5

• Numerical Integration: Intermediate
• Simple Plotting in Scilab
• Notable mistakes in computational physics: spaceflight

  Week 6

• Using Euler's Method to solve a differential equation
• Reading and writing data from and to files

  Week 7

• Using Heun's method to solve a differential equation
• Breaking a program into subroutines, in one file or several files
• In-class exercise: practice putting several functions into one source code file.

  Week 8

• Using integrals of motion as checks on accuracy
• Adaptive timesteps

Additional resources

I've created a set of example Scilab functions which may help you to start writing your own.

I have placed some books on reserve in the RIT Library.

• Code Complete, by Steve McConnell. Everything you wanted to know about coding.
• Writing Solid Code, by Steve Maguire. A succint set of rules to help you avoid common errors and express yourself more clearly.
• The Elements of Programming Style and Software Tools , by Brian Kernighan and P. J. Plauger. Both are old (written in 1974 and 1976, respectively), but filled with excellent, extended samples of Good Code. These books are out of print, but you can find them in the RIT library.
• Engineering and scientific computing with Scilab ed. Claude Gomez. The book is old (published in 1999), but is the only item in the library which uses Scilab as an example language.
• Introduction to Numerical Methods: A MATLAB approach by Abdelwahab Kharab. A book very much like our text, but including a very quick introduction to the MATLAB language. You are very likely to find material pertaining to most of our assignments in this book.

Some links to sites with information and tutorials on Scilab:

• The main Scilab web site
• Getting Started with Scilab a tutorial by Surendran_ks. The link in the previous sentence goes to a local copy of the the gzipped tar file on the main Scilab site
• Scilab Bag of Tricks may explain some arcana
• If you are familiar with MATLAB, you may find it useful to read a comparison of Scilab to MATLAB

Other good references for mathematical techniques:

• Numerical Recipes by Press et al.
• Data Reduction and Error Analysis for the Physical Sciences, by Philip R. Bevington. Published in 1969, but almost always contains the answer to your problem anyway. A classic. If you find it on an "old book" rack, buy it!
• Numerical simulation of physics from the point of view of a computer game writer.
• A molecular dynamics primer by Furio Ercolessi provides some explanation of different algorithms, and is available on-line. The section labelled time integration algorithm is most relevant to this course.
• Molecular Dynamics Simulation by Haile has been recommended as a good starting point, though I haven't read it myself.
• Computer Simulation of Liquids by M.P. Allen and D.J. Tildesley has been recommended as a good reference, though I haven't read it myself.
• A list of numerical methods resources kept at the State University of New York, Stony Brook, has a wealth of links.

You can find other courses on Computational Physics on-line at

• Computational Physics course at Imperial College of London by Angus McKinnon.
• Introduction to Computational Physics taught by Franz J. Vesely at the University of Vienna.
• Physics 780.20: Computational Physics taught by Richard J. Furnstahl at Ohio State University.
• Astronomy 497D, Numerical Methods in Astrophysics, taught by Tom Quinn at the University of Washington.
• Physics 305, Computational Physics II taught at Drexel by Steve McMillan.
• Notes for a course on Numerical Methods put together by Jeremy Schiff. Some of the material is in Hebrew, but much of the material (and references) are in English.
• The Computational Physics course at Simon Fraser University has a complete course manuscript available.

Last modified 3/6/2007 by MWR

Copyright © Michael Richmond. This work is licensed under a Creative Commons License.