Zen and the Art of MatLab

Damian Gordon
 Hard work done by :

Daphne Gilbert
&
Susan Lazarus
 Introduction to MatLab
• MatLab is an interactive, matrix-based
system for numeric computation and
visualisation
• MATrix LABoratory
• Used in image processing, image synthesis,
engineering simulation, etc.
 References
• “Mastering MatLab” Duane Hanselman, Bruce
Littlefield
• “The MatLab Primer”
http://www.fi.uib.no/Fysisk/Teori/KURS/WRK/m
at/mat.html
• “The MatLab FAQ”
http://www.isr.umd.edu/~austin/ence202.d/matlab
-faq.html
Printed Circuit Board
Specific Bond Selected
Bond Shape Estimated
MATLAB Command Window

To get started, type one of these: helpwin, helpdesk, or demo.

For product information, type tour or visit www.mathworks.com.

» help

HELP topics:
 Creating Variables
>> varname = 12
varname =
12

>> SS = 56; N = 4; Tot_Num = SS + N

Tot_Num =
60
 Operations
+ Addition • Add vars
- Subtraction • Subtract vars
* Multiplication Multiplication
^ Power • Raise to the power
\ Division • Divide vars (A div B)
/ Division • Divide vars (B div A)
 Creating Complex Numbers
>> 3 + 2i
>> sqrt(9) + sin(0.5)*j
ans =
3.0000 + 0.4794i
Num = sqrt(9) + sin(0.5)*j
real(Num)
imag(Num)
 Entering Matrices (1)
>> A = [1 2 3; 4 5 6; 7 8 9]

OR
>> A = [
123
456
789]
 Entering Matrices (2)
• To create an NxM zero-filled matrix
>> zeros(N,M)
• To create a NxN zero-filled matrix
>> zeros(N)
• To create an NxM one-filled matrix
>> ones(N,M)
• To create a NxN one-filled matrix
>> ones(N)
 Entering Matrices (3)
• To create an NxM randomly-filled matrix (which
is uniformly distributed)
>> rand(N,M)

• To create an NxM randomly-filled matrix (which

is normally distributed)
>> randn(N,M)
 Complex Matrices
• To enter a complex matrix, you may do it in
one of two ways :
>> A = [1 2; 3 4] + i*[5 6;7 8]

OR

>> A = [1+5i 2+6i; 3+7i 4+8i]

 MATLAB Command Window
» who

Your variables are:

a b c

» whos
Name Size Bytes Class

a 8x8 512 double array

b 9x9 648 double array
c 9x9 648 double array

Grand total is 226 elements using 1808 bytes

 Matrix Addition
» A = [ 1 1 1 ; 2 2 2 ; 3 3 3]
» B = [3 3 3 ; 4 4 4 ; 5 5 5 ]

»A+B
ans =

4 4 4
6 6 6
8 8 8
 Matrix Subtraction
» A = [ 1 1 1 ; 2 2 2 ; 3 3 3]
» B = [3 3 3 ; 4 4 4 ; 5 5 5 ]

»B-A
ans =

2 2 2
2 2 2
2 2 2
 Matrix Multiplication
» A = [ 1 1 1 ; 2 2 2 ; 3 3 3]
» B = [3 3 3 ; 4 4 4 ; 5 5 5 ]

»A*B
ans =
12 12 12
24 24 24
36 36 36
 Matrix - Power
»A^2
ans =
6 6 6
12 12 12
18 18 18

»A^3
ans =
36 36 36
72 72 72
108 108 108
 Matrix Transpose
A=
1 1 1
2 2 2
3 3 3
» A'
ans =
1 2 3
1 2 3
1 2 3
 Matrix Division
Left Division \ Right Division /
x = A\B (is A*x=B) x=A/B (is x*A=B)

>> A = rand(4) >> A = rand(4)

>> B = rand(4) >> B = rand(4)
>> C = A \ B >> C = A / B

=> A * C = B => C * A = B
 Matrix Operations
+ Addition • Add matrices
- Subtraction • Subtract matrices
* Multiplication Matrix Multiplication
^ Power • Raise to the power
‘ Conjugate Transpose • Get transpose
\ Left Division • x = A\B (is A*x=B)
/ Right Division • x=A/B (is x*A=B)
 Philosophiae Naturalis
Principia Matlabematica
Damian Gordon
Getting to Matlab
 Magic Matrix

MAGIC Magic square.

MAGIC(N) is an N-by-N matrix
constructed from the integers
1 through N^2 with equal row, column, and
diagonal sums.
Produces valid magic squares for N =
1,3,4,5,...
 Identity Function
>> eye (4)
ans =

1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
 Upper Triangle Matrix
» a = ones(5) » triu(a)

a= ans =

1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 0 1 1 1 1
1 1 1 1 1 0 0 1 1 1
1 1 1 1 1 0 0 0 1 1
1 1 1 1 1 0 0 0 0 1
 Lower Triangle Matrix
» a = ones(5) » tril(a)

a= ans =

1 1 1 1 1 1 0 0 0 0
1 1 1 1 1 1 1 0 0 0
1 1 1 1 1 1 1 1 0 0
1 1 1 1 1 1 1 1 1 0
1 1 1 1 1 1 1 1 1 1
 Hilbert Matrix
» hilb(4)

ans =

1.0000 0.5000 0.3333 0.2500

0.5000 0.3333 0.2500 0.2000
0.3333 0.2500 0.2000 0.1667
0.2500 0.2000 0.1667 0.1429
 Inverse Hilbert Matrix
» invhilb(4)

ans =

16 -120 240 -140

-120 1200 -2700 1680
240 -2700 6480 -4200
-140 1680 -4200 2800
 Toeplitz matrix.
TOEPLITZ
TOEPLITZ(C,R) is a non-symmetric
Toeplitz matrix having C as its
first column and R as its first row.
TOEPLITZ(R) is a symmetric (or
Hermitian) Toeplitz matrix.

-> See also HANKEL

 Summary of Functions
• magic - magic matrix
• eye(4) - identity matrix
• triu(4) - upper triangle
• tril(4) - lower triangle
• hilb(4) - hilbert matrix
• invhilb(4) - Inverse Hilbert matrix
• toeplitz(4) - non-symmetric
Toeplitz matrix
 Dot Operator
• A = magic(4); b=ones(4);
• A*B
• A.*B
• the dot operator performs element-by-
element operations, for “*”, “\” and “/”
 Concatenation
• To create a large matrix from a group of
smaller ones
• try
– A = magic(3)
– B = [ A, zeros(3,2) ; zeros(2,3), eye(2)]
– C = [A A+32 ; A+48 A+16]
– Try some of your own !!
 Subscripts
• Row i and Column j of matrix A is denoted
by A(i,j)
• A = Magic(4)
• try
– A(1,4) + A(2,4) + A(3,4) + A(4,4)
• try
– A(4,5)
 The Colon Operator (1)
• This is one MatLab’s most important
operators
• 1:10 means the vector
– 1 2 3 4 5 6 7 8 9 10
• 100:-7:50
– 100 93 86 79 72 65 58 51
• 0:pi/4:pi
– 0 0.7854 1.5708 2.3562 3.1416
 The Colon Operator (2)
• The first K elements in the jth column is
– A(1:K, j)
• Sum(A(1:4, 4)) is the sum of the 4th
column
or
• Sum(A(:, 4)) means the same
matlABBA
 The Colon Operator (1)
• This is one MatLab’s most important
operators
• 1:10 means the vector
– 1 2 3 4 5 6 7 8 9 10
• 100:-7:50
– 100 93 86 79 72 65 58 51
• 0:pi/4:pi
– 0 0.7854 1.5708 2.3562 3.1416
 The Colon Operator (2)
• The first K elements in the jth column is
– A(1:K, j)
• Sum(A(1:4, 4)) is the sum of the 4th
column
or
• Sum(A(:, 4)) means the same
 Deleting Rows and Columns (1)
• Create a temporary matrix X
• X=A;
• X(:, 2) = []
• Deleting a single element won’t result in a
matrix, so the following will return an error
• X(1,2) = []
 Deleting Rows and Columns (2)
• However, using a single subscript, you can
delete
– a single element
– sequence of elements
So X(2:2:10) = []
gives
• x = 16 9 2 7 13 12 1
 The ‘FIND’ Command (1)
>> x = -3:3
x=
-3 -2 -1 0 1 2 3

K = find(abs(x) > 1)
K=
1267
 The ‘FIND’ Command (2)
A = [ 1 2 3 ; 4 5 6 ; 7 8 9]

[i, j] = find (A > 5)

i= j=

3 1
3 2
2 3
3 3
 Special Variables
• ans - default name for results
• pi - pi
• eps - “help eps”
• flops - count floating point ops
• inf - Infinity, e.g. 1/0
• NaN - Not a number, e.g. 0/0
• i,j - root minus one
• why - why not ?
LOGO Command
 The ‘PLOT’ Command (1)
>> X = linspace(0, 2*pi, 30);
>> Y = sin(X);
>> plot(X,Y)

>> Z = cos(X);
>> plot(X,Y,X,Z);
 The ‘PLOT’ Command (2)
>> W = [Y ; Z]
>> plot (X,W)

Rotate by 90 degrees
>> plot(W,X)
The Meditations of
Matlab Aurelius
 The ‘PLOT’ Command (1)
>> X = linspace(0, 2*pi, 30);
>> Y = sin(X);
>> plot(X,Y)

>> Z = cos(X);
>> plot(X,Y,X,Z);
 The ‘PLOT’ Command (2)
>> W = [Y ; Z]
>> plot (X,W)

Rotate by 90 degrees
>> plot(W,X)
 PLOT Options
>> plot(X,Y,’g:’)
>> plot(X,Y,’r-’)
>> plot(X,Y,’ko’)

>> plot(X,Y,’g:’,X,Z,’r-’,X,Y,’wo’,X,Z,’c+’);
 PLOT Options
>> grid on
>> grid off
>> xlabel(‘this is the x axis’);
>> ylabel(‘this is the y axis’);
>> title(‘Title of Graph’);
>> text(2.5, 0.7, ’sin(x)’);
>> legend(‘sin(x)’, ‘cos(x)’)
 SUBPLOT Command
• Subplot(m,n,p)
– creates a m-by-n matrix in and plots in the pth
plane.
subplot(2,2,1)
plot(X,Y)
subplot(2,2,2)
plot(X,Z)
subplot(2,2,3)
plot( X,Y,X,Z)
subplot(2,2,4)
plot(W,X)
 Specialised matrices
• compan • Companion matrix
• gallery • Higham test matrices
• hadamard • Hadamard matrix
• hankel • Hankel matrix
• pascal • Pascal matrix.
• rosser • Classic symmetric
eigenvalue test problem
• vander
• Vandermonde matrix
• wilkinson
• Wilkinson's eigenvalue
test matrix
 Polynomials
Polynomials are represented as
row vectors with its coefficients in
descending order, e.g.
X4 - 12X3 + 0X2 +25X + 116

p = [1 -12 0 25 116]
 Polynomials
The roots of a polynomial are found as
follows

r = roots(p)

roots are represented as a column vector

 Polynomials
Generating a polynomial from its roots

polyans = poly(r)

includes imaginary bits due to rounding

mypolyans = real(polyans)
 Polynomial Addition/Sub
• a = [1 2 3 4]
• b = [1 4 9 16]

• c=a+b
• d=b-a
 Polynomial Addition/Sub
• What if two polynomials of different
order ?
X3 + 2X2 +3X + 4
X6 + 6X5 + 20X4 - 52X3 + 81X2 +96X + 84
• a = [1 2 3 4]
• e = [1 6 20 52 81 96 84]
• f = e + [0 0 0 a] or f = e + [zeros(1,3) a]
 Polynomial Multiplication
• a = [1 2 3 4]
• b = [1 4 9 16]
Perform the convolution of two arrays !
• g = conv(a,b)
g=
1 6 20 50 75 84 64
 Polynomial Division
• a = [1 2 3 4]
• g = [1 6 20 50 75 84 64]
Perform the deconvolution of two arrays !
[q,r] = deconv(g, a)
q= {quotient}
1 4 9 16
r= {remainder}
00000000
 Polynomial Differentiation
• f = [1 6 20 48 69 72 44]

• h = polyder(f)

h=
6 30 80 144 138 72
 Polynomial Evaluation
x = linspace(-1,3)

p = [1 4 -7 -10]

v = polyval(p,x)

plot (x,v), title(‘Graph of P’)

 Rational Polynomials
• Rational polynomials, seen in Fourier,
Laplace and Z transforms
• We represent them by their numerator and
denominator polynomials
• we can use residue to perform a partial
fraction expansion
• We can use polyder with two inputs to
differentiate rational polynomials
An Anthropologist on Matlab
 Data Analysis Functions (1)
• corrcoef(x) • Correlation coefficients
• cov(x) • Covariance matrix
• cplxpair(x) • complex conjugate pairs
• cross(x,y) • vector cross product
• cumprod(x) • cumulative prod of cols
• cumsum(x) • cumulative sum of cols
• del2(A) • five-point discrete Laplacian
• diff(x) • diff between elements
• dot(x,y) • vector dot product
• gradient(Z, dx, dy) • approximate gradient
 Data Analysis Functions (2)
• histogram(x) • Histogram or bar chart
• max(x), max(x,y) • max component
• mean(x) • mean of cols
• median(x) • median of cols
• min(x), min(x,y) • minimum component
• prod(x) • product of elems in col
• sort(x) • sort cols (ascending)
• std(x) • standard dev of cols
• subspace(A,B) • angle between subspaces
• sum(x) • sum of elems per col
Symbolic Math Toolbox
 Symbolic Expressions

• ‘1/(2*x^n)’
• cos(x^2) - sin(x^2)
• M = sym(‘[a , b ; c , d]’)
• f = int(‘x^3 / sqrt(1 - x)’, ‘a’, ‘b’)
 Symbolic Expressions
diff(‘cos(x)’)
ans =
-sin(x)

det(M)
ans =
a*d - b * c
 Symbolic Functions
• numden(m) - num & denom of polynomial
• symadd(f,g) - add symbolic polynomials
• symsub(f,g) - sub symbolic polynomials
• symmul(f,g) - mult symbolic polynomials
• symdiv(f,g) - div symbolic polynomials
• sympow(f,’3*x’) - raise f^3
 Advanced Operations
• f = ‘1/(1+x^2)’
• g = ‘sin(x)’

compose(f,g) % f(g(x))
ans=
1/(1+sin(x)^2)
 Advanced Operations
• finverse(x^2)
ans =
x^(1/2)

• symsum(‘(2*n - 1) ^ 2’, 1, ‘n’)

ans =
11/3*n + 8/3-4*(n+1)^2 + 4/3*(n+1)^3
 Symbolic Differentiation
f = ‘a*x^3 + x^2 + b*x - c’
diff(f) % by default wrt x
ans =
3*a^2 + 2*x + b
diff(f, ‘a’) % wrt a
ans =
x^3
diff(f,’a’,2) % double diff wrt a
ans =
0
 Symbolic Integration
f = sin(s+2*x)
int(f)
ans =
-1/2*cos(s+2*x)
int(f,’s’)
ans =
-cos(s+2*x)
int(f, ‘s’, pi/2,pi)
ans=
-cos(s)
 Comments &Punctuation (1)
• All text after a percentage sign (%) is
ignored
>> % this is a comment

• Multiple commands can be placed on one

line separated by commas (,)
>> A = magic(4), B = ones(4), C = eye(4)
 Comments &Punctuation (2)
• A semicolon may be also used, either after a
single command or multiple commands
>> A = magic(4); B = ones(4); C = eye(4);
• Ellipses (…) indicate a statement is
continued on the next line
A = B/…
C
 SAVE Command (1)
>> save
• Store all the variables in binary format in a file
called matlab.mat
>> save fred
• Store all the variables in binary format in a file
called fred.mat
• >> save a b d fred
• Store the variables a, b and d in fred.mat
 SAVE Command (2)
>> save a b d fred -ascii
• Stores the variables a, b and d in a file called
fred.mat in 8-bit ascii format
>> save a b d fred -ascii -double
• Stores the variables a, b and d in a file called
fred.mat in 16-bit ascii format
 Load Command
• Create a text file called mymatrix.dat with
– 16.0 3.0 2.0 13.0
– 5.0 10.0 11.0 8.0
– 9.0 6.0 7.0 12.0
– 4.0 15.0 14.0 1.0

• “load mymatrix.dat”, create variable

mymatrix
 M-Files
• To store your own MatLab commands in a
file, create it as a text file and save it with a
name that ends with “.m”
• So mymatrix.m
A = […
16.0 3.0 2.0 13.0
5.0 10.0 11.0 8.0
9.0 6.0 7.0 12.0
4.0 15.0 14.0 1.0];

• type mymatrix
 IF Condition

if condition If x = 2
{commands} output = ‘x is even’
end end
 WHILE Loop

X = 10;
while condition count = 0;
{commands} while x > 2
end x = x / 2;
count = count +
1;
end
 FOR Loop
for n = 1:10
x(n) = sin(n);
end
for x=array
{commands} A = zeros(5,5); %
end prealloc
for n = 1:5
for m = 5:-1:1
A(n,m) = n^2 + m^2;
end
disp(n)
end
 Creating a function
function a = gcd(a,b)
% GCD Greatest common divisor
% gcd(a,b) is the greatest common divisor of
% the integers a and b, not both zero.
a = round(abs(a)); b = round(abs(b));
if a == 0 & b == 0
error('The gcd is not defined when both numbers are zero')
else
while b ~= 0
r = rem(a,b);
a = b; b = r;
end
end
 Quick Exercise (!)
• Consider Polynomial Addition again :
how would you write a program that takes in two
polynomials and irrespective of their sizes it adds the
polynomials together ? Given that the function length(A)
returns the length of a vector.
Answers on a postcard to : [email protected]

oh, and while you’re here anyhow, if you have a browser

open, please go to the following sites :
http://www.the hungersite.com
http://www.hitsagainsthunger.com
Flash Gordon
and the Mud Men of Matlab
 Quick Exercise (!)
• Consider Polynomial Addition again :
how would you write a program that takes in two
polynomials and irrespective of their sizes it adds the
polynomials together ? Given that the function length(A)
returns the length of a vector.
Answers on a postcard to : [email protected]

oh, and while you’re here anyhow, if you have a browser

open, please go to the following sites :
http://www.the hungersite.com
http://www.hitsagainsthunger.com
 Creating Programs
Title : Program.m

function out = program(inputs)

% PROGRAM
<code>
 Know Thyself
• Where am I ?
– pwd
• Get me onto the hard disk
– cd C:
• Where am I now ?
– pwd
• Get me to where I know
– cd ..
 Quick Answer (!)
function c = mypoly(a,b)
% MYPOLY Add two polynomials of variable lengths
% mypoly(a,b) add the polynomial A to the polynomial
% B, even if they are of different length
%
% Author: Damian Gordon
% Date : 3/5/2001
% Mod'd : x/x/2001
%
c = [zeros(1,length(b) - length(a)) a] + [zeros(1, length(a) -

length(b)) b];
 Recursion
function b = bart(a)
%BART The Bart Simpson program writes on the blackboard
% program, Bart writes the message a few times
% and then goes home to see the Simpsons
if a == 1
disp('I will not....');
else
disp('I will not skateboard in the halls');
bart(a - 1);
end
 Curve Fitting
• What is the best fit ?
– In this case least squares curve fit
• What curve should be used ?
– It depends...
 POLYFIT : Curve Fitting
• polyfit(x,y,n) - fit a polynomial

• x,y - data points describing the curve

• n - polynomial order
• n = 1 -- linear regression
• n = 2 -- quadratic regression
 Curve Fitting Example
x = [0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1];
y = [-.447 1.978 3.28 6.16 7.08 7.34 7.66 9.56 9.48
9.30 11.2];
polyfit(x,y,n)
• n=1
p = 10.3185 1.4400
• n=2
p = -9.8108 20.1293 -0.0317
y = -9.8108x2 + 20.1293x - 0.0317
 Curve Fitting Example
xi = linspace(0,1,100);

z = polyval(p,xi)

plot(x,y,'o',x,y,xi,z,':');
 Interpolation - 1D
• t = interp1(x,y,.75)
t=

9.5200

also
• interp1(x,y,.75,’spline’)
• interp1(x,y,.75,’cubic’)
 Interpolation - 2D
• interp2(x,y,Z,xi,yi,TYPE)

TYPE =

'nearest' - nearest neighbor interpolation

'linear' - bilinear interpolation
'cubic' - bicubic interpolation
'spline' - spline interpolation
 Fourier Functions
• fft • Fast fourier transform
• fft2 • 2-D fft
• ifft • Inverse fft
• ifft2 • 2-D Inverse fft
• filter • Discrete time filter
• filter2 • 2-D discrete tf
• fftshift • shift FFT results so -ve
freqs appear first
 Tensors
• See ‘Programs’

• ‘Tensors’

• ‘Tensors.html’
 3D Graphics
T = 0:pi/50:10*pi;
plot3(sin(t), cos(t), t);
 3D Graphics
title('Helix'), xlabel('sin(t)'),
ylabel('cos(t)'), zlabel('t')
grid
 3D Graphics
• Rotate view by elevation and azimuth

view(az, el);

view(-37.5, 60);