Chap1 Matlab

Contents
1 Programming using Matlab 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Vectors and matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Character strings, cell arrays, structures . . . . . . . . . . . . . . . . 15
1.4 Statement level control structures . . . . . . . . . . . . . . . . . . . . 19
1.4.1 Logical expressions . . . . . . . . . . . . . . . . . . . . . . . . 19
1.4.2 Conditional statements . . . . . . . . . . . . . . . . . . . . . . 23
1.4.3 Loop statements . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.5 Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.5.1 Built-in functions . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.5.2 Function files . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.5.3 Anonymous functions . . . . . . . . . . . . . . . . . . . . . . . 29
1.5.4 Passing functions as input . . . . . . . . . . . . . . . . . . . . 29
1.6 Graphics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.6.1 2-D graphics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.6.2 3-D graphics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.7 Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
1.7.1 Vectorising loops . . . . . . . . . . . . . . . . . . . . . . . . . 37
1.7.2 Preallocating arrays . . . . . . . . . . . . . . . . . . . . . . . 37
i
ii UNSW Sydney, Australia, MATH2301, T1 2024
Chapter 1
Programming using Matlab
1.1 Introduction
In this course, we will use Matlab to carry out all of the programming exercises. Mat-
lab is a software package designed for numerical computation and visualisation. The
symbolic toolbox can do symbolic algebra as well. MATLAB = MATrix LABorartory.
As the name suggests, the building blocks of Matlab are matrices.
Matlab is provides an interactive environment. In a default setup, there are four
main windows: Command Window, Current Directory, Workspace, and Command
History.
• Command Window: This is the main window where all commands including
those for running user-written programs are typed in. The MATLAB command
prompt is >>. E.g.
>> help help

>> help
>> help quad
• Current Directory: This is where all your files from the current directory are
listed. To change directories, click on the menu bar or in the Command Window:
>> cd directory_name
To use a file from another directory, use addpath.
• Workspace: This window lists all variables you have generated in this session,
together with their type and size. Double click on a variable in this window to
view its value. Right click for other options.
1
2 UNSW Sydney, Australia, MATH2301, T1 2024
• Command History: This window records all commands typed in the Command
Window. Double click on a command to execute it again. You can also use the
up-arrow on the keyboard.
A comprehensive set of demonstrations is available by typing the command
>> demo
and following the instructions.

The fundamental data type in MATLAB is the array which includes integers,
real numbers (which are set to be of double precision), matrices, character strings,
structures, and cells.
In constrast to other programming languages, no declarations of data type or data
object are needed. For vectors and arrays, the dimension statements are not required
either.
Matlab stored computed values in variables. You don’t have to declare variable
name as the beginning of the script, but an assignment statement is needed. Variable
names are case-sensitive, can contain up to 31 characters, and must start with a letter.
These are allowable:
My_Variable, my_variable, Math2301.
These are not allowable:
my-variable, 2301math, @UNSW
There is a reserved word list. To check if a variable name var has been used or is in
the reserved list, type in the following command
>> exist(’var’)
The following table lists some special variables: If you assign a new value to one
Variable name Meaning

ans default variable name used for results
pi π
eps smallest number such that when added to 1
gives a number greater than 1 on the computer
inf, Inf ∞
NaN, nan not a number, e.g. √
0/0
i, j imaginary number −1
of the special variables above, its prior value is overwritten and lost.
To clear a variable from the memory, you type in
Matlab 3
>> clear variable_name

To clear all variables from the memory, type in
>> clear all
Sometimes, you don’t want too much information to be printed out on the screen.
A semicolon at the end of a command suppresses the screen output, except for graphics
and on-line help commands. The command more on directs MATLAB to show one
screen of output at a time. To turn it off, type more off. Though MATLAB performs
computations in double precision, the default appearance is a scaled fixed point format
with 5 digits (format short). The command
>> format compact
suppresses blank lines in the output, thus allowing more information to be displayed
on the screen.
MATLAB provides its own built-in editor with an integrated debugger. You can
use this editor to create and save your own programs in files. Alternatively, you can
use any text editor, like vi or emacs. MATLAB has 3 types of files:
• M-files: standard ASCII text files, with a .m extension to the file name. There
are 2 types of M-files: script files and function files.
• Mat-files: binary data files with a .mat extension to the file name.
• Mex-files: MATLAB callable Fortran and C programs, with a .mex extension

to the file name.
A MATLAB script is a text file containing MATLAB commands, and having a
.m extensions. To execute commands in a script, simply type the name of the file,
without the .m extension, at the MATLAB prompt. Running a script file is equivalent
to typing all the commands (in that file) one by one at the MATLAB prompt. The
advantage is obvious: script files are easy to modify. The script file has to be in the
current directory, or in a directory of the MATLAB search path. The command what
can be used to check if the file is in the current directory. A path can be added by
the command addpath. As a warning, the name of a script file cannot be the same
as the name of a variable it computes.
Example 1.1.1 Write a MATLAB script, named area.m, to compute the areas of
the circles of radii π 1/3 and π 2/3 . On running the script should print out the following
line (with no other outputs on the screen)
Radius Area
. . . . . .
. . . . . .
where ... contains the appropriate value.
% MATLAB script to compute the areas of circles of

% radii pi^(1/3) and pi^(2/3).
disp([’ Radius Area’])
disp(’ ’)
r = pi^(1/3); A = pi*r^2;
disp([r, A])
r = pi^(2/3); A = pi*r^2;
disp([r, A])
To execute, type in the MATLAB command window: area

If some variables are defined, they can be saved for next sessions, using the com-
mand save filename. A file named filename.mat is created in the current directory.
The command load filename reloads the values in the next session.
Example 1.1.2 Try the following commands on the command window.
>> a = pi; b = a^2; c = a+b;

>> save a_value a % save value of a only
>> save ab_value a b % save value of a and b
>> save all_values % save all values in session
>> clear all
>> a, b, c % What do you expect to see?
>> load ab_value % Do not include .mat
>> a, b, c % What do you expect to see?
>> load all_values
>> c
For reading and writing formatted binary and text files, the following commands
should suffice fopen, fprintf, fscanf, and fclose.
One of the following commands opens a file named MyFile.
% Open file named MyFile for read access.
% File must exist before command is executed
>> fid = fopen(’MyFile’);
% or
>> fid = fopen(’MyFile’,’r’);
% Open or create file for writing,
% existing contents overwritten
Matlab 5
>> fid = fopen(’MyFile’,’w’);

% Open or create file for writing,
% append data to end of file
>> fid = fopen(’MyFile’,’a’);
% Open file for reading and writing
>> fid = fopen(’MyFile’,’r+’);
% Open or create file for reading and writing
>> fid = fopen(’MyFile’,’w+’);
% Open or create file for reading and appending
>> fid = fopen(’MyFile’,’a+’);
The file identifier fid (an integer) is used as the first argument to fprintf and
fclose. It has the value −1 when a file cannot be opened.
>> x = [1:10];
>> y = sqrt(x);
>> fid = fopen(’square_root’,’w’);
>> fprintf(fid,’ Square Root\n ’);
>> fprintf(fid,’\n’);
>> fprintf(fid,’ %3i %8.2f\n’,[x;y]);
>> fclose(fid);
To read and write binary data to a file, use fread and fwrite.
1.2 Vectors and matrices

A row vector is defined by listing numbers which are separated by either commas or
spaces. A column vector has entries separated by semicolons or new lines.
>> v1 = [ 1 2, 3 4 ]
v1 =
1 2 3 4
>> v2 = [ 3 ; 4
6 ]
v2 =
3
4
6
>> v1(3), v2(2)
ans =
3
ans =
4
Row and column vectors are one-dimensional arrays. Matrices are two-dimensional
arrays. It is not necessary to declare the dimension of a matrix.
>> A = [ 1 2 3
4 5 6 ]
A =
1 2 3
4 5 6
>> B = [ 3 7 9 ; 2 6 4 ]
B =
3 7 9
2 6 4
If it is not possible to type the entire input on the same line, use an ellipsis (three
dots) and continue the input on the next line.
>> A = [ 4 7 9 7 ...
8 9 ]
A =
4 7 9 7 8 9
The size command gives the number of rows and columns of an array. The length
command gives the maximum value of these two numbers.
>> A = [1 8 0 3 ; 2 4 9 6 ; 0 3 7 9];
>> size(A)
ans =
3 4
>> length(A)
ans =
4
>> [m n] = size(A) % assign these values to variables
% m and n for later access
m =
3
n =
4
>> size(A,1)
ans =
3
>> size(A,2)
ans =
4
Matlab 7
>> a = [2 3 4 5 9];
>> size(a)
ans =
1 5
>> length(a)
ans =
5
>> B = [] % empty array
B =
[]
>> size(B)
ans =
0 0
>> a(3)
ans =
4
>> a(2) = 0
>> a
a =
2 0 4 5 9
>> A(2,3), A(5)
ans =
9
ans =
4
>> A(3,1) = 8
A =
1 8 0 3
2 4 9 6
8 3 7 9
Example 1.2.1 Solve the equation Ax = b where

   
1 8 3 2
A =  2 4 6  and b =  3 
8 3 9 5
>> A = [1 8 3 ; 2 4 6 ; 8 3 9]; b = [2 ; 3 ; 5];

>> x = inv(A)*b
x =
0.1500
0.0833
0.3944
>> x = A\b % Almost the same as inv(A)*b but

x = % faster and more numerically stable
0.1500
0.0833
0.3944
Notes:
>> c = [1 3 4];
>> x = A\c
??? Error using ==> mldivide
Matrix dimensions must agree.
>> x = A\c’
x =
-0.1500
-0.0833
0.6056
Given a scalar a (real or complex) and two matrices A and B. The following opera-
tions are available under some appropriate conditions.
Command Meaning Condition

C =a+A Cij = a + Aij
C =a−A Cij = a − Aij
C =a∗A Cij = a ∗ Aij
C = A/a Cij = Aij /a
C =A+B Cij = Aij + Bij size(A)=size(B)
C =A−B Pij − Bij
Cij = A size(A)=size(B)
C =A∗B Cij = nk=1 Aik Bkj n =size(A,2)=size(B,1)
C = Abn C = |A ∗ ·{z
· · ∗ A} size(A,1)=size(A,2)
n times
C = A′ Cij = Aji
C = inv(A) C = A−1 size(A,1)=size(A,2)
C = A/B C = AB −1 size(A,2)=size(B,1)=size(B,2)
C = A\B C = A−1 B size(A,1)=size(A,2)=size(B,1)
Consider the following commands
>> v = [3:9] % increment by 1

v =
3 4 5 6 7 8 9
>> v = 3:2:9 % increment by 2
v =
Matlab 9
3 5 7 9
>> v = [3:2:8]
v =
3 5 7 % final value does not exceed 8
>> v = [8:-2:3] % negative increment
v =
8 6 4 % final value does not go below 3
>> v = [5:2]
v =
Empty matrix: 1-by-0
>> v = 2:-1:3
v =
The command linspace(a,b,n) generates a vector of n linearly equally spaced
points between (and including) a and b. If n is not specified, i.e. linspace(a,b), the
default value is n = 100.
Example 1.2.2 Generate a vector of 5 equally spaced points between and including
1 and 10.
>> linspace(1,10,5)
ans =
1.0000 3.2500 5.5000 7.7500 10.0000
This is the same as
>> [1:(10-1)/(5-1):10] % [a:(b-a)/(n-1):b]
ans =
1.0000 3.2500 5.5000 7.7500 10.0000
If a is a scalar, and A and B are two arrays of the same size, then the element-
by-element operations give the following arrays:
Operation Meaning
C = a./A Cij = a/Aij
C = A.\a Cij = a/Aij
C = A.ˆa Cij = Aaij
C = a.ˆA Cij = aAij
C = A. ∗ B Cij = Aij Bij
C = A./B Cij = Aij /Bij
C = A.\B Cij = Bij /Aij
B
C = A.ˆB Cij = Aijij
>> v = 1:5
v =
1 2 3 4 5
>> 2.^v
ans =
2 4 8 16 32
>> v.^2
ans =
1 4 9 16 25
>> v^2
??? Error using ==> mpower
Matrix must be square.
To access parts of a matrix, consider the following commands:
>> A = [2 5 -1 7 ; 8 0 6 3 ; 9 8 -2 1 ; 3 7 9 2]
A =
2 5 -1 7
8 0 6 3
9 8 -2 1
3 7 9 2
>> B = A(2:4,3:4)
B =
6 3
-2 1
9 2
>> C = A(1:2:4,:)
C =
2 5 -1 7
9 8 -2 1
>> A(2:4,3:4) = 0
A =
2 5 -1 7
8 0 0 0
9 8 0 0
3 7 0 0
>> A(1:2:end,1) = -1
A =
-1 5 -1 7
8 0 0 0
-1 8 0 0
3 7 0 0
Using an array to refer to part of a matrix:
Matlab 11
>> j = [1 3 4];
>> A(3,j)
ans =
-1 0 0
We can stack small matrices one on top of each others.
>> A = [1 -1 ; 2 2]; B = [2 1 ; 0 0]; C = [1:4 ; 5:8];

>> E = [A B ; C]
E =
1 -1 2 1
2 2 0 0
1 2 3 4
5 6 7 8
>> D = [1:5, 5:8];
>> F = [A B ; D]
??? Error using ==> vertcat
All rows in the bracketed expression must have the
same number of columns.
Zeros (or ones) arrays are arrays that contains all zeros (or all ones).
>> zeros(3)
ans =
0 0 0
0 0 0
0 0 0
>> zeros(2,3)
ans =
0 0 0
0 0 0
>> A = [1 2 5 7 ; 2 4 8 1];
>> zeros(size(A))
ans =
0 0 0 0
0 0 0 0
Replacing zeros by ones gives similar matrices with entries 1.

We can define identity matrices as follows
>> eye(2)
ans =
1 0
0 1
>> eye(3,2)
ans =
1 0
0 1
0 0
>> eye(2,3)
ans =
1 0 0
0 1 0
>> A = [1 2 5 7 ; 2 4 8 1];
>> B = eye(size(A))
B =
1 0 0 0
0 1 0 0
We can access diagonals of a matrix by the diag command. If A is a matrix then
diag(A,k), where k is an integer, produces a column vector whose components are
entries on the kth diagonal of A.
>> A = [1 3 5 ; 2 4 6]; v1 = diag(A), v2 = diag(A,1), ...
v3 = diag(A,2), v4 = diag(A,3), v5 = diag(A,-1)
v1 =
1
4
v2 =
3
6
v3 =
5
v4 =
v5 =
2
The same command can be use to create diagonal matrices. If v is a vector, then
diag(v,k) creates a square matrix of appropriate size whose kth diagonal is v.
>> v = [1:3]; v1 = diag(v), v2 = diag(v,1), v3 = diag(v,-2)
v1 =
1 0 0
0 2 0
0 0 3
v2 =
Matlab 13
0 1 0 0
0 0 2 0
0 0 0 3
0 0 0 0
v3 =
0 0 0 0 0
0 0 0 0 0
1 0 0 0 0
0 2 0 0 0
0 0 3 0 0
Sparse matrices are matrices which contain a lot of zeros.
>> A = [0 0 1 0 0 ; -2 0 3 0 0 ; 0 0 -5 0 0]
A =
0 0 1 0 0
-2 0 3 0 0
0 0 -5 0 0
>> nnz(A) % number of non-zero entries of A
ans =
4
>> B = sparse(A)
B =
(2,1) -2
(1,3) 1
(2,3) 3
(3,3) -5
The command full(B) reproduces a matrix identical to A. Arithmetic and index-
ing operations can be applied to sparse matrices or to a mixture of sparse and full
matrices.
>> clear all % clear all existing variables from memory

>> clc % clear the Command Window
>> A = [2 0 5 0 0 ; 0 0 0 4 0 ; 0 0 0 1 0]; B = sparse(A);
>> nnz(A); v = [9 11 8];
>> who
Your variables are:
A B ans v
>> whos A
Name Size Bytes Class
A 3x5 120 double array
Grand total is 15 elements using 120 bytes
>> whos

B 3x5 72 double array (sparse)
ans 1x1 8 double array
v 1x3 24 double array
>>
>> A = eye(200); B = sparse(A);
>> whos
B 200x200 3204 double array (sparse)
For more infomation on how sparse matrices are stored, read J.R. Gilbert, C. Moler,
and R. Schreiber, Sparse matrices in MATLAB: Design and implementation. http:
//www.mathworks.com/help/pdf_doc/otherdocs/simax.pdf.
Example 1.2.3 A sparse matrix B of size 4 × 4 with nonzeros entries B(2, 1) = −2,
B(1, 2) = 1, B(4, 2) = 1, B(2, 3) = 3, B(3, 3) = −5, B(4, 4) = 1 can be created as
follows:
>> a = [-2 1 1 3 -5 1]; % non-zero entries of B

>> r = [2 1 4 2 3 4]; % row indices
>> c = [1 2 2 3 3 4]; % column indices
>> B = sparse(r,c,a,4,4)
B =
(2,1) -2
(1,2) 1
(4,2) 1
(2,3) 3
(3,3) -5
(4,4) 1
If A is a given matrix, the command sA = sparse(A) produces a matrix sA which

is a sparse version of A.
Sparse matrices are naturally defined in terms of their diagonals using spdiags
command.
>> e = ones(5,1); % some none-zero entries

>> A = spdiags([-e 2*e -e],... % 0 -> diagonal, -1 -> subdiagonal
[-1 0 1],5,5); % 1 -> superdiagonal
>> full(A)
Matlab 15
ans =
2 -1 0 0 0
-1 2 -1 0 0
0 -1 2 -1 0
0 0 -1 2 -1
0 0 0 -1 2
1.3 Character strings, cell arrays, structures

In MATLAB a string is an array of ASCII character codes. (ASCII = American
Standard Code for Information Interchange).
>> s = ’Hello world!’
s =
Hello world!
>> whos s
s 1x12 24 char array
>> isstr(s), isstr(5), isstr(’5’) % check if the variable
ans = % is a string
1
ans =
0
ans =
1
Since strings are arrays, they can be manipulated with all array manipulation tools.
>> a = ’Hello’; b = ’world’; c = ’!’;
>> [a b c]
ans =
Helloworld!
>> [a ’ ’ b c]
ans =
Hello world!
>> a(1:4)
ans =
Hell
>> b(end:-1:1)
ans =
dlrow
>> u = ’I don’’t know how’ % single quotes within character are
u = % symbolised y two consecutive quotes

I don’t know how
Strings can have multiple rows, but each row must have equal number of columns.
>> v = [’Joshua’ ; ’Jack ’ ; ’Ben ’]
v =
Joshua
Jack
Ben
Artificial creation of blank spaces can be avoided by using char command.
>> student = char(’Joshua’, ’Jack’, ’Ben’)
student =
Joshua
Jack
Ben
Horizontal concatenation of string arrays having the same number of rows is accom-
plished by the function strcat.
>> strcat(’Adam’, ’ Smith’)
ans =
Adam Smith
>> students = char(’Joshua’, ’Jack’, ’Ben’);
>> surnames = char(’ Waugh’, ’ Warne’, ’ Bevan’);
>> v = strcat(students, surnames)
v =
Joshua Waugh
Jack Warne
Ben Bevan
>> whos v
v 3x12 72 char array
A cell array is an array that can contain any MATLAB object, e.g., an array of
numbers, logicals, a string , etc.
>> A = {[1 2 ; 3 4], ’MATH2301’, 1 + 2i, 1 > 0}
A =
[2x2 double] ’MATH2301’ [1.0000 + 2.0000i] [1]
>> whos A
A 1x4 305 cell array
Matlab 17
The curly braces indicate that the expression respresents a cell rather than, e.g., a
numerical value. We can preallocate a variable as a cell array, using the command
cell.
>> B = cell(2,2)
B =
[] []
[] []
>> B{1,1} = ones(1,3); B{1,2} = rand(1,2);
>> B{2,1} = ’MATLAB’; B{2,2} = A
B =
[1x3 double] [1x2 double]
’MATLAB’ {1x4 cell }
To display the content of a cell array, use celldisp.

>> celldisp(A)
A{1} =
1 2
3 4
A{2} =
MATH2301
A{3} =
1.0000 + 2.0000i
A{4} =
1
>> celldisp(B)
B{1,1} =
1 1 1
B{2,1} =
MATLAB
B{1,2} =
0.8913 0.7621
B{2,2}{1} =
1 2
3 4
B{2,2}{2} =
MATH2301
B{2,2}{3} =
1.0000 + 2.0000i
B{2,2}{4} =
1
To display the content of a component of a cell array:
>> A{1,2}
ans =
MATH2301
>> B{2,2}
ans =
[2x2 double] ’MATH2301’ [1.0000 + 2.0000i] [1]
>> B{2,2}{2}
ans =
MATH2301
>> B{2,2}{1}
ans =
1 2
3 4
>> B{2,2}{1}(2,2)
ans =
4
A structure is a collection of objects labelled by field names. E.g.:
>> s = struct(’name’, ’harry’, ’age’, 45)
s =
name: ’harry’
age: 45
>> whos s
s 1x1 266 struct array
>> s.name, s.age
ans =
harry
ans =
45
>> fieldnames(s)
ans =
’name’
’age’
We can also form array of structures:
>> t = struct(’name’, {’harry’, ’sue’, ’john’}, ...
’age’, {46, 37, 29})
t =
1x3 struct array with fields:
name
Matlab 19
age
>> t(2).name
ans =
sue
>> t.age
ans =
46
ans =
37
ans =
29
>> disp([t.age])
46 37 29
1.4 Statement level control structures

1.4.1 Logical expressions
Logical expressions are statements that involve logical variables, relational operators
and logical operators. The logical variables take value 1 (true) or 0 (false) in
MATLAB. A logical variable can be created by using:
• The true, false functions.
• The logical function: converts numeric values to logical.
• Relational operators.
• Logical operators.
• All is..." functions. E.g. isinteger, isempty, isnan, isinf, etc.
The use of true/false functions is listed below.
>> x = [true false]

x =
1 0
>> A = true(2), B = false(2,3)
A =
1 1
1 1
B =
0 0 0
0 0 0
>> whos x A B
A 2x2 4 logical array
B 2x3 6 logical array
x 1x2 2 logical array
The use of logical function is listed below.
>> a = 1;
>> b = logical(a); % convert a numeric value to logical
>> whos a b
a 1x1 8 double array
b 1x1 1 logical array
>> A = [1 2 0 ; -4 0 3 ];
>> B = logical(A)
Warning: Values other than 0 or 1 converted to logical 1.
B =
1 1 0
1 0 1
>> whos A B
B 2x3 6 logical array
The following table list the relational operators supported by MATLAB.
Operator Description
== equal
e= not equal
< less than
<= less than or equal
> greater than
>= greater than or equal
>> 3 > 4
ans =
0 % 0 when relation is false
>> s = 2 < 3 % assign the answer to variable s.
s =
Matlab 21
1 % 1 when relation is true

>> u = [1 3 9 0]; v = [2 1 9 3]; % u and v same size
>> s1 = u ~= v, s2 = u >= v
s1 =
1 1 0 1
s2 =
0 1 1 0
>> A = [2 3 ; 4 5]; B = [2 4 ; 5 5]; % A and B same size
>> C = A ~= B
C =
0 1
1 0
>> whos s1 s2 C
C 2x2 4 logical array
s1 1x4 4 logical array
s2 1x4 4 logical array
The following table list all logical operators. Note: A, B, C are arrays of same
size.
Operator Description (
1, A(i, j) ̸= 0 and B(i, j) ̸= 0
& C = A&B, C(i, j) =
0, otherwise
(
1, A(i, j) ̸= 0 or B(i, j) ̸= 0
| C = A|B, C(i, j) =
0, otherwise
(
1, A(i, j) = 0
e C = eA, C(i, j) =
0, otherwise
(
1, at least one u(i) ̸= 0
any v = any(u), v =
0, otherwise
(
1, all u(i) ̸= 0
all v = all(u), v =
0, otherwise.
>> A = [2 0 3 ; 5 0 1]; B = [0 0 -3 ; 1 3 0];

>> C = A & B, D = A | B, E = ~A,
C =
0 0 1
1 0 0
D =
1 0 1
1 1 1
E =
0 1 0
0 1 0
>> u = [2 0 1];
>> v1 = any(u)
v1 =
1
>> v2 = all(B) % v2 = [all(B(:,1)) all(B(:,2)) all(B(:,3))]
v2 =
0 0 0
The is... functions: isequal, isempty, isfinite, isinf, isinteger, isprime,isnan

. . .
>> x1 = []; x2 = [1 2];
>> v1 = isempty(x1), v2 = isempty(x2)
v1 =
1
v2 =
0
>> A = [2 0 3 ; 5 0 1]; B = [0 -3 ; 1 3]; C = [2 0 3; 5 1 1];
>> v1 = isequal(A,A), v2 = isequal(A,B), v3 = isequal(A,C)
v1 =
1
v2 =
0
v3 =
0
Logical arrays are used in array indexing and in conditional statements if, elseif.
Example 1.4.1 Find all elements of x which are < 5 and ≥ 0. Set new value 5 to
those elements.
>> x = [5 2 8 4 9 0 -1];
>> v = x < 5 & x >= 0
v =
0 1 0 1 0 1 0
>> x(v)
ans =
2 4 0
>> x(v) = 5 % same as x(x<5&x>=0) = 5
Matlab 23
x =
5 5 8 5 9 5 -1
The zero values cannot be used to index a vector or a matrix.
>> v = [0 1 0 1 0 1 0];
>> x(v)
??? Subscript indices must either be real positive
integers or logicals.
>> A = [2 0 3 ; 5 0 1]; B = [0 0 -3 ; 1 3 0];
>> A(A&B) = inf
A =
2 0 Inf
Inf 0 1
>> A = magic(3), A(~isprime(A)) = 0
A =
8 1 6
3 5 7
4 9 2
A =
0 0 0
3 5 7
0 0 2
1.4.2 Conditional statements

Conditional statements or if statements provide a mechanism for branching in a
computer program, i.e. they make it possible to execute selectively on the basis of
information generated while the program is running. In MATLAB, an if statement
takes the following form:
if <logical expression 1>

<statement 1>
elseif <logical expression 2>
<statement 2>
else
<statement 3>
end
Zero or more than one elseif branch is permitted. The else branch can also be omitted.
Example 1.4.2 Let x be given. For each element of x, if it is nonpositve then negate
it, otherwise just square it.
>> x = [-1 2 5 -3];

>> if x <= 0
y = -x;
else
y = x.^2;
>> end
>> disp([x;y])
-1 2 5 -3
1 4 25 9
1.4.3 Loop statements

A loop is a programming construct in which a group of statements is executed re-
peatedly. In MATLAB, loops are of two kinds: for loops and while loops. A for
loop has the form
for n = 1:N % a value for N has to be pre-defined
<statement>
end
Example 1.4.3 The following commands
>> for n = 1:5
disp([num2str(n) ’^2 = ’ num2str(n^2)])
end
will produce
1^2 = 1
2^2 = 4
3^2 = 9
4^2 = 16
5^2 = 25
Note: the command num2str converts numbers to a string.
Example 1.4.4 To find the unique whole number K such that (K − 1)2 ≤ 1073 <
K 2 , we could use the following script:
for n = 1:40
if n^2 > 1073
break
end
end
disp([’K = ’ num2str(n)])
disp([num2str(n-1) ’^2 = ’ num2str((n-1)^2)])
Matlab 25
Example 1.4.5 What is the output of the following commands?

>> A = [8 1 6 ; 3 5 7; 4 9 2];
>> for J = 1:3
for I = 1:3
if (I == J)
A(I,J) = 2;
elseif abs(I-J) == 1
A(I,J) = -1;
else
A(I,J) = 0;
end
end
end
>> A
A while loop has the form
while <logical expression>
<statement>
end
The for loop in the last example could be replaced with the following while loop:
n = 1;
while n^2 < 1073
n = n + 1;
end
disp([’K = ’ num2str(n)])
disp([num2str(n-1) ’^2 = ’ num2str((n-1)^2)])
Obviously, the statements in a while loop must modify the value of the logical
expression, so that the latter eventually becomes false and the loop terminates. There
is always the danger of an infinite loop, in which the logical expression never becomes
false, but is always true, perhaps due to programmer not considering all possibilities.
The for ... break is safer.
1.5 Functions
1.5.1 Built-in functions
MATLAB has numerous built-in functions. Use the command help elfun to see
availabe elementary mathematical functions. Scalar functions naturally take scalar
arguments; but will also act on matrices elementwise.
>> x = 1.2; y = [1:0.2:2]; X = [1 2 ; pi 5];

>> sin(x), sin(y), sin(X)
ans =
0.9320
ans =
0.8415 0.9320 0.9854 0.9996 0.9738 0.9093
ans =
0.8415 0.9093
0.0000 -0.9589
Vector functions act on vectors and matrices.
Example 1.5.1 If v is a row vector or column vector, sum(v) gives the sum of the
elements of v. The command sum(v,1) treats v as a matrix and adds its components
columnwise. The command sum(v,2) finds the sum of elements in each row.
>> v = [1:5];
>> s1 = sum(v), s2 = sum(v,1), s3 = sum(v,2)
s1 =
15
s2 =
1 2 3 4 5
s3 =
15
The commands prod(v), prod(v,1), and prod(v,2) have the same meaning for
products of elements of v.
Matrix functions take matrix arguments. For example, given a square matrix A,
the command det(A) computes the determinant of A, inv(A) computes the inverse
of A, eig(A) computes the eigenvalues of A. Let B be a 4 × 3 matrix, the command
reshape(B,2,6) will take elements of B columnwise and form a new 2 × 6 matrix.
>> A = [1 3 5 ; 2 4 6 ; 9 2 1];
>> det(A)
ans =
-12
>> inv(A)
ans =
0.6667 -0.5833 0.1667
-4.3333 3.6667 -0.3333
2.6667 -2.0833 0.1667
>> eig(A)
ans =
10.8674
Matlab 27
-5.0846
0.2172
>> B = [1 3 5 ; 2 4 6 ; 9 2 1 ; 4 8 2];
>> reshape(B,2,6)
ans =
1 9 3 2 5 1
2 4 4 8 6 2
1.5.2 Function files

Function files (which are M-files) allow one to create one’s own MATLAB commands.
A function file usually has these lines:
function [yout1, yout2, ...] = funcname(xin1, xin2, ...)
% comment line 1
% comment line 2
% ...
yout1 = ...;
yout2 = ...;
...
Notes:
• The function M-file name and the function name that appears in the first line
of the file should be identical. In reality, MATLAB ignores the function name
in the first line and executes functions based on the file name.
• All comment lines immediately following the function definition line are dis-
played by help funcname.
• A single output does not require []. E.g. function y = funcname(xin1,xin2).
• Input variable names given in the function definition line are local to the func-
tion. An input variable name can be the name of another function. On execu-
tion, this name must be passed on as a character string, i.e., enclosed within
two single right quotes.
• All variables inside a function are local and are erased after execution of the
function, whereas those in a script file are left in the MATLAB workspace after
execution of the script.
• Functions can have arguments, script files do not.
• Inside the body of a user-defined function, nargin and nargout return the
numbers of declared input and output arguments. nargin(’funcname’) and
nargout(’funcname’) return those values for the M-file function funcname.m.
Example 1.5.2 Write a MATLAB function defining the following mathematical

function: 

 0 for x < −2
(x + 2)2 /4 for − 2 ≤ x < 0

f (x) =


 (3 − x)/3 for 0 ≤ x < 3
for x ≥ 3

0
function y = piecefunc(x)
% EXAMPLE of a piecewise-defined function
%
% (x+2)^2/4 for -2 <= x < 0,
% y = (3-x)/3 for 0 <= x < 3,
% 0 otherwise
y= zeros(size(x)); % preallocate y
i1 = find(-2 <= x & x < 0)
i2 = find( 0 <= x & x < 3)
y(i1) = (x(i1)+2).^2/4;
y(i2) = (3-x(i2))/3;
Example 1.5.3 Write a function file to define the probability density function
−(x − µ)2

1
f (x) = √ exp .
2πσ 2σ 2
function density = normal_dist(x,mu,sigma)

% function density = normal_dist(x,mu,sigma)
% computes the probability density function for a normally
% distributed random variable with mean mu and standard
% deviation sigma. If the argument mu and sigma are omitted,
% then a standard normal distribution is assumed
% (mu = 0 and sigma = 1).
%
if nargin == 1
mu = 0;
sigma = 1;
end
density = exp(-0.5* ((x-mu)/sigma).^2)/(sqrt(2*pi)*sigma);
The following script file will illustrate the use of normal dist.
% Script file exec_normal_dist.m

% using the function normal_dist, with 1 and 3 input
% arguments
Matlab 29
x = linspace(-4,4);
y1 = normal_dist(x);
y2 = normal_dist(x,1,0.5);
plot(x,y1,x,y2)
legend(’standard normal’,’mu=1, sigma=0.5’)
trapz(x,y1) % trapezoidal rule to evaluate integral
trapz(x,y2) % type "help trapz" to learn more
1.5.3 Anonymous functions

We can also define an anonymous function either at the MATLAB command line:
>> fun1 = @(x) x.^2;

>> fun2 = @(x,y) x+y;
or in an M-file (script file or function M-file):
sqr = @(x) x.^2;

x = linspace(0,2);
plot(x,sqr(x));
1.5.4 Passing functions as input

A function handle is a MATLAB value that provides a mean of referring to a function
indirectly. For example, if
sqr = @(x) x.^2
then sqr is the handle of the anonymous function x2 . If a function is defined in a

function M-file, then a handle to this function can be defined by
funhandle = @filename.
Some MATLAB built-in functions (e.g. fzero and feval) or user-defined func-
tions have as input another function. The function handle is used to pass on this
input.
With the function sqr defined above we have
>> feval(sqr,2) % This is the same as sqr(2)

ans =
4
For functions defined in function M-files like piecefunc in Example 1.5.2, there are
several ways to pass it on as an input of another function.
>> piecefunc_handle = @piecefunc;

>> feval(piecefunc_handle,2)
ans =
0.3333
>> feval(@piecefunc,2)
ans =
0.3333
>> feval(’piecefunc’,2)
ans =
0.3333
1.6 Graphics
1.6.1 2-D graphics
Example 1.6.1 Plot y = exp(x) for 0 ≤ x ≤ 3 using the plot command.
% create some points on the domain [0,3]

x = 0:0.3:3;
% compute function values at x
y = exp(x);
plot(x,y)
% We may need more points to get a smooth plot
% create 100 equally spaced points on [0,3]
x = linspace(0,3);
y = x.^2 .* exp(x);
plot(x,y)
Different linestyles (+ * -), and colours (y m c r g b w k) can be chosen in the

options of the plot command. Type help plot for more information.
>> x = linspace(0,1); y = log(x); plot(x,y,’g+’)
The fplot command selects points on the x-axis automatically.

fplot(’sin(10*x)’, [0 3])
fplot(’x^2 * exp(x)’, [-1,1])
% or
fplot(’x.^2 .* exp(x)’,[-1,1])
% Change axes. Type help axis for more info
axis([0 0.5 0 1])
The following commands add labels to the axes, title and legend to the plot: xlabel,
ylabel, title, text, legend. Use help for more information.
Matlab 31
Example 1.6.2 Plot the Bessel functions of order zero, one, and two on [0, 10].
(Bessel functions of order µ are solutions of the ODE x2 y ′′ + xy ′ + (x2 − µ2 )y = 0.
Type help besselj to get more info.)
% This script plots 3 Bessel functions of order 0,1,2 on

% [0,10]. Type "help besselj" for more info.
close all
figure(2)
x = [0:0.2:10];
y0 = besselj(0,x); y1 = besselj(1,x); y2 = besselj(2,x);
plot(x,y0,x,y1,x,y2)
legend(’\mu=0’,’\mu=1’,’\mu=2’) % \mu to produce Greek letter mu
xlabel(’x’)
ylabel(’B_\mu(x)’) % underscore _ to produce subscript
title(’Bessel functions B_\mu’)
Another way to plot multiple graphs in one figure window is to use hold on:
>> x = linspace(0,2); y1 = x.^2; y2 = sin(x);
>> plot(x,y1)
>> hold on
>> plot(x,y2)
>> hold off
To save a figure: use print
>> print -dps filename % save as .ps file
>> print -deps filename % save as .eps file
>> print -depsc filename % save as .eps file (colour)
>> print -djpeg filename % save as .jpeg file
Note: MATLAB 2014b seems to have some problems with exporting Greek sym-
bols to eps files. The quick fix is to use the following commands (taking the previous
plot as an example)
>> title(’Bessel function $B_\mu$’,’interpreter’,’latex’);
>> print -depsc2 filename
Example 1.6.3 Plot sin(x) and cos(x) in one figure for a ≤ x ≤ b, where a and b
are input parameters.
%See file plot_sincos.m

a = input(’a = ’); % prompt for user input
b = input(’b = ’); % prompt for user input
if a >= b
error(’a must be strictly less than b’)

end
x = [a:0.1:b];
y1 = sin(x);
y2 = cos(x);
plot(x,y1,x,y2)
legend(’sin x’, ’cos x’)
print -deps sincos
The error built-in function displays the message, exits the M-file, and returns
control to the keyboard.
PN
Example 1.6.4 Plot the Riemann zeta function ζ = n=1 n−s for s ∈ [a, b] and
a > 1.
disp(’Plot zeta(s) for a < s < b where a > 1’)

a = input (’a = ’);
b = input (’b = ’);
if a <= 1
error(’a must be strictly greater than 1’)
elseif b <= a
error(’b must be strictly greater than a’)
end
s = linspace(a,b);
N = input(’Number of terms = ’);
n = [1:N];
mat1 = repmat(n’,1,length(s));
mat2 = -repmat(s,length(n),1);
y = sum(mat1 .^ mat2);
plot(s,y)
xlabel(’s’), ylabel(’zeta(s)’);
In the previous example, we have used the repmat command. The actions of the
command can be seen in the following code
>> a=2; b=4; s=linspace(2,4,3);

>> N=5; n = [1:N];
>> mat1 = repmat(n’,1,length(s))
mat1 =
1 1 1
2 2 2
3 3 3
4 4 4
Matlab 33
5 5 5
>> mat2 = -repmat(s,length(n),1)
mat2 =
-2 -3 -4
-2 -3 -4
-2 -3 -4
-2 -3 -4
-2 -3 -4
Parametric curves can be drawn as follows:
>> t = linspace(-pi,pi);
>> plot(cos(t),sin(3*t))
>> xlabel(’x(t)’) % x-axis label
>> ylabel(’y(t)’) % y-axis label
>> title(’Parametric curve (cos(t),sin(3t))’)
Curves can be drawn using polar coordinates:
>> theta = linspace(-pi,pi,300);

>> r = 1 ./ (1+0.6*cos(10*theta));
>> polar(theta,r)
Sometimes, we need to put many plots in the same figure. The command subplot
will help us to achieve this. The figure window can be divided into rectangular panes
that are numbered rowwise. Plots are then output to each pane.
x = linspace(0,pi);
y1 = sin(x);
y2 = x.^2 ./ (1+x.^3); % note the dots
subplot(1,2,1) % divide the graphic window into 2 panes,
% 1 row and 2 columns, and get the figure
% in the first pane
plot(x,y1) % plot in the first pane
subplot(1,2,2) % get the figure in the 2nd pane
plot(x,y2) % plot in the second pane
1.6.2 3-D graphics

Type help graph3d to obtain a list of functions available for 3D graphics. Colors can
be used to represent the 4th dimension or values of functions defined on surfaces. The
most commonly used commands for 3D graphics are: plot3, mesh, surf. Graphics
can be rotated by using the Rotate 3D button on the figure menu, or using the view
command. Type help view for more details.
Example 1.6.5 Plot the circular helix x = cos t, y = sin t, z = t for t ∈ [0, 7π].
>> t=linspace(0,7*pi,1000);
>> plot3(cos(t),sin(t),t);
>> axis on
>> grid on
To plot a surface in R3 , firstly we need to create a grid via the meshgrid command.
>> x = -2:2; % create the vector x = [-2 -1 0 1 2]
>> y = -1:1; % create the vector y = [-1 0 1]
>> [X,Y] = meshgrid(x,y) % create a grid of 15 points
% their x-coordinates are stored in X
% their y-coordinates are stored in Y
X =
-2 -1 0 1 2
-2 -1 0 1 2
-2 -1 0 1 2
Matlab 35
Y =
-1 -1 -1 -1 -1
0 0 0 0 0
1 1 1 1 1
In general X, Y are matrices of size n × m where m and n are the lengths of x and y,
resp, and X(i, j) = x(j), Y (i, j) = y(i).
In the following example, we will plot surfaces in 3D using mesh and surf com-
mands.
Example 1.6.6 Plot the surface z = 1−x2 /3−y 2 /5 for −2 ≤ x ≤ 2 and −1 ≤ y ≤ 1.

The surface is determined by points (X(i, j), Y (i, j), Z(i, j)) = (x(j), y(i), Z(i, j)).
>> % see plotS.m

>> x = linspace(-2,2,20);
>> y = linspace(-1,1,20);
>> [X,Y] = meshgrid(x,y);
>> Z = 1 - X.^2/3 - Y.^2/5;
>> subplot(1,2,1)
>> mesh(X,Y,Z) % or mesh(x,y,Z)
>> subplot(1,2,2)
>> surf(X,Y,Z) % or surf(x,y,Z)
Matlab 37
1.7 Efficiency
Performance of your MATLAB code can be improved by vectorising loops and preal-
locating arrays.
1.7.1 Vectorising loops

MATLAB is a matrix language designed for vector and matrix operations. To speed
up your M-file code you should use vectorising algorithms that take advantage of this
design. Whenever possible, convert for and while loops to vector or matrix operations.
Test the following code on your system. In the code, cputime is used to compute
the difference in time. cputime is a built-in function that returns the CPU time (in
seconds) that has been used by the MATLAB process since MATLAB started.
Here is one way to compute the sine of 100001 values ranging from 0 to 10:
t0 = cputime;
i = 0;
for t = 0:.01:1000
i = i + 1;
y(i) = sin(t);
end
t1 = cputime;
loop_time = t1-t0
The vectorised version of the same code is
t0 = cputime;
t = 0:.01:1000;
y = sin(t);
t1 = cputime;
vec_time = t1-t0
1.7.2 Preallocating arrays

A loop that incrementally increases the size of a data structure each time through the
loop can adversely affect performance and memory use. Repeatedly resizing arrays
often requires that MATLAB spend extra time looking for larger contiguous blocks
of memory and then moving the array into those blocks. If you cannot vectorise a
piece of code, you can make your loops go faster by preallocating any vector in which
output results are stored.
Compare the following two pieces of code, assuming that a value for n is given in
advance.
% Without preallocation
clear x
for j=1:n
for i=1:n
x(i,j)=(i+j)/n;
end
end
% With preallocation
clear x
x=zeros(n,n); % preallocate x
for j=1:n
for i=1:n
x(i,j)=(i+j)/n;
end
end

Chap1 Matlab

Uploaded by

Copyright:

Available Formats

Chap1 Matlab

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chap1 Matlab

Uploaded by

Copyright:

Available Formats

Contents

1 Programming using Matlab 1

Programming using Matlab

>> help help

To use a file from another directory, use addpath.

A comprehensive set of demonstrations is available by typing the command

and following the instructions.

My_Variable, my_variable, Math2301.

These are not allowable:

my-variable, 2301math, @UNSW

Variable name Meaning

>> clear variable_name

• Mex-files: MATLAB callable Fortran and C programs, with a .mex extension

where ... contains the appropriate value.

% MATLAB script to compute the areas of circles of

To execute, type in the MATLAB command window: area

Example 1.1.2 Try the following commands on the command window.

>> a = pi; b = a^2; c = a+b;

>> fid = fopen(’MyFile’,’w’);

1.2 Vectors and matrices

Example 1.2.1 Solve the equation Ax = b where

>> A = [1 8 3 ; 2 4 6 ; 8 3 9]; b = [2 ; 3 ; 5];

>> x = A\b % Almost the same as inv(A)*b but

Command Meaning Condition

Consider the following commands

>> v = [3:9] % increment by 1

We can stack small matrices one on top of each others.

>> A = [1 -1 ; 2 2]; B = [2 1 ; 0 0]; C = [1:4 ; 5:8];

Replacing zeros by ones gives similar matrices with entries 1.

>> clear all % clear all existing variables from memory

Name Size Bytes Class

>> a = [-2 1 1 3 -5 1]; % non-zero entries of B

If A is a given matrix, the command sA = sparse(A) produces a matrix sA which

>> e = ones(5,1); % some none-zero entries

1.3 Character strings, cell arrays, structures

u = % symbolised y two consecutive quotes

To display the content of a cell array, use celldisp.

1.4 Statement level control structures

• The true, false functions.

• The logical function: converts numeric values to logical.

• All is..." functions. E.g. isinteger, isempty, isnan, isinf, etc.

The use of true/false functions is listed below.

>> x = [true false]

1 % 1 when relation is true

>> A = [2 0 3 ; 5 0 1]; B = [0 0 -3 ; 1 3 0];

The is... functions: isequal, isempty, isfinite, isinf, isinteger, isprime,isnan

The zero values cannot be used to index a vector or a matrix.

1.4.2 Conditional statements

if <logical expression 1>

>> x = [-1 2 5 -3];

1.4.3 Loop statements

Example 1.4.5 What is the output of the following commands?

>> x = 1.2; y = [1:0.2:2]; X = [1 2 ; pi 5];

1.5.2 Function files

Example 1.5.2 Write a MATLAB function defining the following mathematical

function density = normal_dist(x,mu,sigma)

% Script file exec_normal_dist.m

1.5.3 Anonymous functions

>> fun1 = @(x) x.^2;