- Here is a link to Songxu, the TA for Math 464.
- Syllabus(dvi)
- Computer Projects(dvi)
- Numerical Recipes in Fortran. This is a source book for Fortran routines. Many of them can be easily changed into Matlab functions.
- Numerical Recipes in C is the C version of this book.
- For those of you that are interested in using maple, there are worksheets made up by Greg Arden. They are in the files ~arden/public/{ worksheet1.ms, worksheet2.ms }, which you may copy. Call up maple from a menu or x-window and select file and then open from a maple menu. Then choose one of the worksheet files to open. You will be prompted through the worksheet.
- Michael Scholz has informed me that the Engineering Library Copy Center sells a copy of the Matlab Primer for $1.50. I suspect it is the older edition of the Primer, but you might find it adequate.
- Frequently Asked Questions (dvi). (Udated 11/30/95)
- There will be a colloquium in the Mathematics Department on Tuesday, October 17 that many of you will be interested in. Yves Nievergelt will talk about accelerating convergence of numerical methods. The talk will be at 4:00 pm in Padelford C-36. Yves promises the talk will be at the level of 308 and 464 students.
- (10/13/95) The terminology that I used in class on Wednesday, October 11 might have confused some of you. I referred to the diagonal row above the main diagonal as the upper diagonal. In the computer project this diagonal is referred to as the superdiagonal. Simlarly the lower diagonal is the subdiagonal.
- (10/17/95) Suppose an iterative method is written in the form x=Mx+b. Then we have proved that if ||M||<1 for some norm then the iteration converges. However this doesn't mean that if ||M||>1 for a particular norm then the iteration diverges. This is a delicate point. We have only discussed a few of the infinitely many possible matrix norms. It is not easy to give necessary and sufficient conditions for the convergence of an iteration. I will supply more details later.
- (10/24/95) A good reference for matrix iteration is Matrix Iterative Analysis by Richard Varga (QA 263 v3).
- (10/24/95) An old edition of the Matlab Primer can be viewed from this link. I am not sure if you can bring it up from computers outside the Math Department.
- (10/25/95) There will be an introduction to Matlab lecture given by Randy Leveque today (10/25/95) in Sieg 323, 3-4 pm. It might be of interest to some of you. He will show some numerical analysis examples.
- (10/26/95) Sample problems for the midterm (Wednesday, November 1). You will be allowed one notebook sized (8.5" x 11") page of notes on the midterm. You may write on only one side of the page.
- (10/30/95) Pete Stewart has made his Afternotes on Numerical Analysis available for public perusal. A first glance at them indicates they are quite good.
- (10/31/95) The Computing Seminar will meet in Sieg 323 Wednesday, November 1 at 3pm. Tom Milac will give an introduction to Mathematica.
- (10/31/95) There is a misprint in computer project #2. You should run your program on the systems in part a) and b) of computer project 1.
- (11/2/95) You may use the matlab functions for norms in your computer projects.
- (11/13/95) There will be an undergraduate talk by Professor Michael Eastwood on Drawing with Complex Numbers on Wednesday, November 15 at 4:00 pm in Smith 120. Here is the abstract for more details.
- (11/15/95) If you want to write a matlab routine that takes a function as an argument, you include in the argument list a variable that represents the function name. When you call the function, you should supply the name of the function inside quote marks. For example:
function y=newt(f,df,x,n)
% newton's method newt(f,df,x,n)
% x is the initial guess; n is the number of iterations
% files f.m and df.m contain the function and its derivatives
% the function and derivative names f and df must be supplied in
% quotes, i.e. a call would be of the form, newt('sin', 'cos', .2, 4).
Your program will need to use feval to evaluate the functions.
- (11/15/95) Here is a message about the Computing Seminar today (11/15/95).
Today's Computing Seminar will meet in Sieg 323 at 3pm. The topic is
PV-Wave, a graphics package available on the Applied Math computing system.
I don't think it's available elsewhere on campus, so this may be of interest
primarily to AMath students.
- Randy LeVeque
- (11/21/95) Here is the latest schedule for the computing seminar.
This week in the Computing Seminar Tom Milac will talk about Fortran ---
compiling, loading, and running codes
linking in subroutine libraries
Makefiles
etc.
Next week Wendy Welch will discuss Totalview, a window-based debugging
package for Fortran, C, and C++. Examples from Fortran and C will be used
to demonstrate the main features of this debugger.
The seminar meets Wednesday 3-4pm in Sieg 323.
- (11/28/95) An algorithm for nested multiplication to evaluate the Newton form of an interpolating polynomial is given in the homework problem, #10, p. 223 of Johnson and Riess.
- (11/28/95)
This week in the Computing Seminar Wendy Welch will discuss Totalview,
a window-based debugging package for Fortran, C, and C++.
Examples from Fortran and C will be used
to demonstrate the main features of this debugger.
The seminar meets 3-4pm in Sieg 323.
- (11/29/95) (Lecture correction) The first version of Simpson's rule I used should have had the form ((b-a))/6)[f(a)+4f((a+b)/2)+f(b)]. Then when I made the computation of the error I used h=(b-a)/2, so the form should have been (h/3)[f(a)+4f((a+b)/2)+f(b)]. I should have written h=(b-a)/2 and then there would have been no confusion.
- (11/30/95) For the computer project, you CANNOT assume that the interpolation points are equally spaced (they are not in part (d)). Hence you must use divided differences and not forward differences to compute the interpolating polynomial. This means that you cannot use problem #10 directly but must modify it for your program. If you consult the faq file you will see how to nest the evaluation of the polynomial. You must write a routine to do this.
- (12/5/95) Sample(dvi) problems for the the final, which is at 8:30 am on Wednesday, December 13 in the classroom.
- (12/5/95) There will be a brief intro to latex in the Computing Seminar on 12/6 at 3pm in Sieg 323. Basic formats, equations, labels, figures (from postscript) and bibliographies will be covered.
- (12/7/95) There is a colloquium today that might interest you.
Title: The Mathematicians Meet Escher: The Local Theorem for Tilings
Speaker: Professor Doris Schattschneider, Moravian College
Abstract: The property of a tiling being isohedral is a global one:
the symmetry group of the whole tiling must act transitively on the tiles.
Yet Escher's isohedral tilings were made according to a local rule: every tile
must be surrounded in the same way. The Local Theorem (for tilings in any
n-dimensional space) gives a charcterization of isohedral tilings that tells
us Escher had the right idea.
PLACE: CMU 326
DATE: Thursday, Thursday, Dec. 7, 1995
TIME: 4:00 p.m. -- 5:00 p.m.
TEA: 3:30 in the Mathematics Lounge
- (12/8/95) Review (dvi) topics.
- (12/8/95) I will have office hours on Monday, December 11, and Tuesday, December 12, from 9:30-11:30.
- (12/11/95) The final will be at 8:30 am on Wednesday, December 13 in the classroom.