
MME 523
Analysis For Applications
Fall 2003
instructor:
John Goulet (goulet@wpi.edu)
Kaven Hall, room 202
Purposes of course:
Each class will:
Misc Files:
roundoff error
(9/9)
Outline
of Topics
1. Root Findi
bisection, Newton’s Method
differentials
here
is the demo from class on Sept 2: (excel file)
Newton's Method in Maple
(.mws Maple file)
2. Integration
Riemann Sums
Trapezoidal follow this
link for the first assignment (9/23)
Simpson’s Rule
3. Series and Sequences
including
power series
follow
this link for the assignment for
follow
this link for the assignment for
great
animation showing Taylor Series convergence!
4. Regression Analysis
October 21 assignment and notes
November 4 assignment
and the products
of that assignment
5.
The Hunt for Pi Nov 11
more stuff about Pi than
you ever wanted to know!
6. Differential Equations (1st
order)
7.
Putting things together
Assignments:
#1
for Sept 9, please
·
get an Academic computer account
·
bring a floppy disk
·
install Maple at home
·
read through Maple
website looking for essential appropriate to your needs
see especially this
link for both help and downloads
·
work on
Bisection and Newton’s in both
Excel and Maple
·
get a calculus book to serve as a reference (any one will do)
·
Maple capabilities needed
o
set up a function f:=x
-> x^3 + 2;
o
assignment x:=3;
o
loops while (b-a > e) do
stuff od;
o
plotting plot(
f(x), x = 2..5);
o
convert to decimal
evalf(%);
#2 (Sept 16, for Sept 23)
follow
this link.
A
good reference on
Now let’s try solving some problems!! In each case, we would like to get the answer
to 4 decimal places with Bisection .
Please make a note of how many iterations in each it takes to achieve
this. Let’s use Excel for Bisection and Maple for Newton’s. We will work on Newton’s Method as well as
Maple on 9/9 .
1) x3 + 2x – 6 = 0 ; the root
in [1,2]
2) x4 + 5x3 + 1 = 0 ; the root in [-1, 0 ]
3) 7x3 + x – 5 = 0 (the only root, please)
4) the root of
cos(x) = 2x
5) x – 2 + 2 ln(x) = 0
A question lurking in
the background is this: if you want 4
decimal place accuracy, how small does the
eventual range for the root have to be??? In other words, how do you translate “decimal
places” into
“tolerance” ???