Matlab Databook

Tim Love
July 24, 2006

Abstract
This document is aimed at 3rd and 4th years who need to revise matlab for
use in coursework. Many of the examples are taken from Prof J.M.Maciejowski’s
Getting started with Matlab and Dr Paul Smith’s Tutorial Guide to Matlab. most
of the material also applies to using octave. Suggestions and contributions for
this document are welcomed. Mail [email protected].

Contents
1 How to avoid needing this document 2

2 Starting Matlab 3

3 Finding your way around matlab 3

3.1 Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.2 Functions and scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.3 Directories and the path . . . . . . . . . . . . . . . . . . . . . . . . . . 3

4 Matrix operations 4
4.1 Creating Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.2 Picking sub-matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.3 Matrix operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.4 Matrix functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

5 Maths 5
5.1 Complex Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.2 Numerical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.3 Sparse Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.4 Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
5.5 Symbolic Maths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

6 Screen textual output 7

7 I/O 7
7.1 Saving and loading data . . . . . . . . . . . . . . . . . . . . . . . . . . 7
7.2 Saving graphics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
7.3 Printing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

c
Copyright 2004-6 by T.P.Love. This document may be copied freely for the pur-
poses of education and non-commercial research. Cambridge University Engineer-
ing Department, Cambridge CB2 1PZ, England.

1
8 Graphics 8
8.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8.2 3D plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8.3 Plotting experimental data . . . . . . . . . . . . . . . . . . . . . . . . . 9
8.4 Movies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

9 Graphics and Objects 9

10 Programming Constructs 10
10.1 if . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
10.2 while . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
10.3 for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
10.4 switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

11 Writing your own scripts and functions 11

11.1 Global variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

12 Debugging 12

13 Miscellaneous 12
13.1 Stopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
13.2 Program-generated names and commands . . . . . . . . . . . . . . . 12
13.3 Vectorising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
13.4 Calling Fortran or C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

1 How to avoid needing this document

Matlab has online tutorials and easily accessible lists of routines. Also CUED keeps
a list of articles written here and elsewhere. So consider the following web links
before reading further, especially if you’re reading this document on paper.

• Getting Started1
• Using MATLAB2
• Mathematics3

• Graphics4
• 3-D Visualisation5
• MATLAB Functions Listed by Category6
• MATLAB Functions Listed Alphabetically7
• CUED’s matlab page8
1 http://www.mathworks.com/access/helpdesk/help/techdoc/learn matlab/learn matlab.shtml
2
http://www.mathworks.com/access/helpdesk/help/techdoc/intro tutorial.html
3 http://www.mathworks.com/access/helpdesk/help/techdoc/math anal/math anal.shtml
4 http://www.mathworks.com/access/helpdesk/help/techdoc/creating plots/creating plots.shtml
5 http://www.mathworks.com/access/helpdesk/help/techdoc/visualize/visualize.shtml
6 http://www.mathworks.com/access/helpdesk/help/techdoc/ref/ref.shtml
7 http://www.mathworks.com/access/helpdesk/help/techdoc/ref/refbookl.shtml
8 http://www-h.eng.cam.ac.uk/help/tpl/programs/matlab.html

2
 • Reading the online version of this document9 saves you having to type ex-
amples in - you can just copy/paste. You won’t have to type the WWW
addresses either! A PDF version10 is online if you need a paper copy.

2 Starting Matlab
On the teaching system you can type matlab at a Terminal window, or look in the
Programs/ Matlab submenu of the Start menu at the bottom-left of the screen.
Depending on the set-up, you might start with several windows (showing docu-
mentation, etc) or just a command line. This document will focus on command
line operations.

3 Finding your way around matlab

3.1 Variables
• To list the variables currently accessible, use who (or for more details, whos).
The variables will be arrays. To find the size of an array A use size(A), which
will return the number of rows and columns.
• To remove a variable A use clear A.

3.2 Functions and scripts

Matlab has many functions available
• If you know the name of a function but you don’t know how to use it, try the
help command - e.g. help sum tell you how to sum the elements of an array.
Throughout this document details of commands will be omitted - using help
or, better still, the WWW documents will help you find all there is to know.
• If you don’t know the name of the function you need, look through the lists
in the first section, or use the lookfor command - e.g. lookfor sum will tell
you the names of all the functions that might have something to do with
summing.

3.3 Directories and the path

When you type a command, matlab has to decide how to respond. First it checks
to see if it has a built in function of that name. Then it checks through a list of
directories looking for an appropriately named file.
• To see what’s actually going to be run when you type a command, use which
- e.g. which sum shows that the sum command is built into matlab.
• To see what directory matlab’s in, type pwd. To change directories use the cd
command.
• what lists the matlab-specific files in the current directory.
9 http://www-h.eng.cam.ac.uk/help/tpl/programs/Matlab/matlabdatabook/
10
http://www-h.eng.cam.ac.uk/help/tpl/programs/Matlab/matlabdatabook/matlabdatabook.pdf

3
 • To see which directories are searched, use path. Most of the directories in the
path belong to matlab. The final place it looks is the current directory. If you
have a subdirectory called matlab in your home directory, matlab will look
there before looking in any other directory. The ordering of these directories
matters - if you have a file in the current directory that’s the same name as a
system file, the system’s file will be run rather than yours.

4 Matrix operations
4.1 Creating Matrices
A 1x1 matrix can be created using something like A=7;. A=7:10; creates a matrix
with 4 elements, namely 7, 8, 9 and 10. A=7:0.5:10; creates a matrix with element
values running from 7 to 10 with a step-size of 0.5.
Between [ and ] commas (or spaces) are used to put things in rows, while
semicolons (or line-breaks) put things in columns. So
b = [ 1 3 5; 2 4 6];
creates a matrix with 2 rows and 3 columns.
Matrices can contain strings - e.g. A=’bat’.
There are commands to create particular types of matrices - zeros, ones, eye,
diag, rand, etc.

4.2 Picking sub-matrices

Rows, columns and submatrices can be selected from a matrix. Here are some
examples. Note how parentheses (rather than square brackets) are used when ma-
nipulating matrices (the % signs introduce comments)
b = [ 1 3 5; 2 4 6];
c=b(2,:) % row 2
d=b(:,1) % column 1
e=b(:,2:3) % columns 2 and 3
Matrices can be composed from submatrices, so f=[d e] would produce a matrix
equal to the original b.

4.3 Matrix operators

Matrix operators work much as expected. In order of precedence they are listed in
table 1.
’ transpose
^ power
∗ multiplication
/ right division
\ left division
+ addition
− subtraction

Table 1: Matrix operators

Note that the normal matrix rules apply, so matrices need to be appropriately
shaped. Using .* (note the extra dot) multiples corresponding elements together.

4
./ and .\ work in a similar way. .^ raises each element to the given power (rather
than the matrix as a whole).

4.4 Matrix functions

Functions exist to perform common operations on matrices. A few are listed in
table 2

inv inverse
det determinant
trace trace
eig eigenvectors and eigenvalues
rank estimate rank
null calculate a basis for the null space
rref Gaussian elimination on an augmented matrix
lu LU decomposition
qr QR decompostion
svd SVD
pinv Pseudoinverse
rot90 rotation of 90 degrees
triu extract upper triangle
tril extract lower triangle

Table 2: Matrix functions

Some functions are especially useful when used with string matrices. E,g.
A=’abcdefg’
A=fliplr(A)
A=sort(A)
circshift(A,[0 1])

5 Maths
As well as the usual operators there are many functions. See table 3

5.1 Complex Numbers

√
By default i and j have the value −1 and complex numbers can be represented
in the usual way - e.g s = 2i+3. The usual operators and functions can be used
on complex numbers and there are some special functions too - imag, real, conj,
angle, etc. Note that abs works with complex numbers, and that matrix transpo-
sition by default conjugates the elements (use .’ rather than ’ to suppress this).

5.2 Numerical Analysis

It helps to know some numerical analysis terms.
Inf :- The IEEE representation of positive infinity.
NaN :- The IEEE representation of ‘Not-a-Number’, used when an operation has
an undefined result (e.g. 0/0). All logical operations involving NaN return
false, except for 6=. If you try to plot points with the value NaN, nothing
appears. This can be useful for clipping, making holes in surfaces etc.

5
 cos cos (angle needs to be in radians)
sin sin (angle needs to be in radians)
tan tan (angle needs to be in radians)
exp exponential
log natural log
log10 base 10 log
sinh hyperbolic sin
cosh hyperbolic cos
tanh hyperbolic tan
acos inverse cos
acosh inverse hyperbolic cos
asin inverse sin
asinh inverse hyperbolic sin
atan inverse tan
atan2 2-argument form of inverse tan
atanh inverse hyperbolic tan
abs absolute value
sign sign (either -1 or +1)
round round to the nearest integer
floor round down to the nearest integer
ceil round up to the nearest integer
fix round towards 0
rem remainder after integer division
sqrt square root

Table 3: Math functions

Condition Number :- The condition number of a matrix (calculated using cond)

gives a fair indication of how well the matrices can be operated upon.

5.3 Sparse Matrices

If you know that your matrices will stay nearly empty, then make them sparse -
they’ll take up less space and can be operated on more quickly.

% comparison of sparse vs full matrices

% Create sparse and full versions of the same matrix.
% Note that the density is 0.05. The break-even point
% for this operation seems to be about 0.25.
S = sprandn(60,60,0.05);
F = full(S);
% Compare speed.
disp(’Sparse’)
t0=cputime;
B = S * S;
cputime-t0
disp(’Full’)
t0=cputime;
C = F * F;
cputime-t0

6
5.4 Algebra
Matlab stores polynomials as vectors. For example, 2x 2 − 3x + 1 can be represented
as p=[2 -3 1] and roots found by using roots(p). poly and polyval are also
useful.

5.5 Symbolic Maths

Toolboxes are adds-ons. One of the most useful and underused (the Symbolic Math
toolbox) performs factorisation, integration, etc, so int(’1/(1+x^2)’) works, as
does
dsolve(’D2y = -a^2*y’, ’y(0) = 1’, ’Dy(pi/a) = 0’)
which solves an ODE. The toolbox also makes function plotting easier - ezplot(’sin(3*t)’,[-10
10]) becomes possible.

6 Screen textual output

• The format command controls how numbers are output. For example format
long g will display 15 digits in either exponential or fixed format, which ever
it thinks best.
• Finishing a command or variable initialisation with a semi-colon suppresses
the output,

• You can display the contents of an array just by typing its name (without
a semi-colon) but that will display the array name too. Particularly if the
array’s a string the disp command is useful - it suppresses the printing of
the array name.

7 I/O
7.1 Saving and loading data
By default matlab saves values into MAT-files with a .mat extension, so save results
Z would save the Z array into a file called results.mat. This array could be
reloaded into matlab by using load results.
You can load files produced by other programs. If a file called data contains “1
2 3” then load data will create a matlab matrix called data which contains those
numbers.

7.2 Saving graphics

If you want to save graphics in a particular format (postscript, jpeg, tiff, PNG,
etc) use the print command. For later inclusion into a LaTeX document, print
-depsc (colour encapsulated postscript) might be best. For Word documents, try
print -depsc -tiff, which adds a TIFF preview.

7.3 Printing
Using print or the print menu option will print the current figure to the default
printer. On the teaching system the cuedpr command will do this in a paper-
efficient way.

7
8 Graphics
8.1 General
Before you display graphics you have to calculate coordinates - unlike some other
programs you can’t just print sin(x) and give a range. Typically you’d do some-
thing like the following
x=linspace(0, pi, 100);
y=sin(x);
plot(x,y);
The linspace command creates a vector of 100 elements equally spaced between
0 and pi inclusive.
Once you have a graph it’s easy to add titles - use the menus on the graphics
window or type commands like
title(’Trig’);
xlabel(’Radians’);
ylabel(’Sin’);
You can plot more than one function on a graph. If you use hold on before
doing another plot, the first plot won’t be erased. Alternatively, you can give the
plot command 2 lines at once. The advantage of this approach is that the lines are
automatically drawn in different colours and that legends are easily created - e.g.
x=linspace(0, pi, 100);
y=sin(x);
y2=cos(x);
plot(x,y,x,y2);
legend(’Sin’,’Cos’);
Adding a grid is easy - use grid on. The menu options let you change many
features of graphs, but you can also use the axis command to change many axis
features - range, aspect ratio, etc.
The plot command has many options to control colour, markers, etc. There’s
also scatter, bar, pie, stairs, stem, polar, compass, etc. Type help graph2d or
(if you want to see some graphics!) see the Graphics tutorial11
Note that you can have many figure windows open (read about the figure
command). You can also have many graphs in each figure. The subplot command
creates axes in tiled positions. So for example if you want 3 rows each containing
2 graphs, and you want to plot vector x in the bottom left graph (the 4th graph,
counting left-to-right), you type
subplot(3,2,4);
plot(x);
Use clf to clear the current figure.

8.2 3D plots
There are many 3D options too, but again you still need to calculate all the coordi-
nates.
11
http://www.mathworks.com/access/helpdesk/help/techdoc/creating plots/intro ne.html

8
 x=linspace(-1, 1, 10);
y=linspace(-1, 1, 10);
[X, Y]=meshgrid(x,y);
Z= X.^2-Y.^2;
mesh(Z);
The meshgrid command creates 2 2D matrices, each with as many elements as
there are grid-points. The X matrix has the x coordinates and the Y matrix the
y coordinates. Then the z coordinate is calculated for each point and put in a Z
matrix. Instead of mesh, surf could be used to display a facetted surface. Then the
following can be done to add graduated colour and a key.
shading interp;
colorbar

8.3 Plotting experimental data

Most of matlab’s 3D facilities expect the data to be evenly spaced on the XY plane,
but what if you have some scattered experimental data? First you need to inter-
polate the data. Suppose you have your datapoints in 3 arrays: x, y, and z. Then
the following will extrapolate the data onto a coarse 10 by 10 grid and display a
colour-coded surface.
xi=linspace(min(x), max(x), 10);
yi=linspace(min(y), max(y), 10);
[X, Y]=meshgrid(xi,yi);
Z=griddate(x,y,z,XI,YI)
surf(Z);
shading interp;
colorbar
There’s also contour, scatter3, waterfall, slice and ezplot. The colormap
command controls palettes. Type help graph3d for details.

8.4 Movies
To create an animated sequence you first need to draw and save each frame. This
example creates a 16-frame movie then runs it 5 times.
% movie demo
M=moviein(16);
for j = 1:16
plot(fft(eye(j+16)))
M(:,j)= getframe;
end
movie(M,5)

9 Graphics and Objects

Though it’s not essential to know how matlab graphics are organised, it’s likely to
be useful sooner or later.
Figures are made of objects (lines, text, images etc). Objects have properties
(like Color, Position, etc) which can be individually controlled. Each object has a
handle (a unique identifier). To experiment with handles, create a simple graph by
typing ‘plot(1:3)’. Typing ‘gcf’ returns the handle of the current figure. ‘get(handle )’

9
lists the properties of the given object, so get(gcf) lists all the properties of the
current figure. You can use set to change the values of these properties once you
know the handle of the object and name of the property. For example, set(gcf,’Color’,’green’)
sets the color of the current figure to green. If you type set(gcf) you get a list of
the properties of the current figure, and alternative settings for these properties.
The axes are children of the figure - and consequently inherit some of the fig-
ure’s properties. If you type get(gcf,’Children’) you’ll get the axes handle,
which you could also get by typing ‘gca’. The axes in turn can have children.
‘set’ is a powerful mechanism. Often matlab has commands that spare you
needing to use set directly, but particularly if you want to retrospectively fix some-
thing, the facility is useful.

10 Programming Constructs
Matlab is a language, not just a calculator. It has the usual programming con-
structs. Both if and while understand all the usual boolean operations (though
note that ~ rather than ! is used for negation - see table 4).

== equal
~= not equal
> greater than
>= greater than or equal
< less than
<= less than or equal
& logical AND
| logical OR
~ logical NOT

Table 4: Matrix functions

10.1 if
if (all(abs(t) >0),
lt = log(t);
else
disp(’There are zeroes in t’);
end

10.2 while
k=1;
while t(k) == 0,
k=k+1;
end
disp(’First nonzero entry in t is’, disp(t(k))

10.3 for
This command is often not needed because matlab functions (sin for example)
often work on whole arrays. The loop below adds an element to back on each
iteration.

10
back=[];
for k=10: -1 : 1,
back=[back,k]
end

10.4 switch
This is like a multiple if statement.
sport=’tennis’;
switch sport
case ’soccer’
disp(’you like soccer’)
case ’tennis’
disp(’you like tennis’)
otherwise
disp(’maybe you like some other sport’)
end

11 Writing your own scripts and functions

It soon gets tedious to type everything on the command line. If you put any valid
sequence of Matlab statements into a file (using an editor) and give the file a name
ending with the suffix .m, then Matlab will execute those statements if you type
the name of the file (without the suffix). For example, if you type edit dothis,
then matlab will start an editor ready to edit a file called dothis.m. If you save the
following text into it
back = []
for k=10:-1:1, back=[back,k], end
then just typing dothis will cause the for loop to be executed. Such files are called
script files.
If the first line of such a file has the form
function outputs = somename(inputs)
then it becomes a function. In this case inputs and outputs are formal parameters,
and variables used inside the file will be local to that file, which means that they
will not clash with other variables you may have used having the same names
(unless these are explicitly declared to be global).
If you wanted to calculate centres of gravity repeatedly, then you could write a
function which would perform the required calculation, and store it for future use.
All you need to do is to create a file called cofg.m which contains the following text:
function cg = cofg(weights, locations)
% Comments are introduced by percent signs, as here.
% These initial comments will be written to the terminal
% if you type ’help cofg’. It is useful to remind the
% user how to call the function by putting in the line:
% cg = cofg(weights, locations)

cg = weights * locations’ / sum(weights);

and that’s all there is to it! When this file exists in your directory, you can type the
following in Matlab:

11
cg = cofg([3 1 7],[-2 0 5])
Of course it may be advisable to make your function more sophisticated, for exam-
ple to check whether both input arguments are row vectors, or to make sure that it
will work with either row or column vectors.
It is also possible to pick up subroutines which have been written in other lan-
guages and pre-compiled. See the Interface Guide12 for details.

11.1 Global variables

Ordinarily, each function has its own local variables, which are separate from those
of other functions, and from those of the base workspace. However, if several func-
tions, and possibly the base workspace, all declare a particular name as global,
then they all share a single copy of that variable. Any assignment to that variable,
in any function, is available to all the other functions then declare it as global.

12 Debugging
• Use size, whos and which to check on variables (or selectively remove semi-
colons from commands so that output isn’t supressed).
• Be careful when choosing variable names
– i begins with the value sqrt(-1) but typing i=7 will change its value.
– If you give a variable the same name as a function odd things can hap-
pen. E.g. the following causes an error on the last line
A=3
size(A)
size = 1
size(A)

• The keyboard command stops function/script execution so you can examine

and change values. To continue, type RETURN.
• Sometimes the error in your program triggers an error message from a func-
tion that your program directly or indirectly calls. Use depfun to see the
dependent functions.
• On some systems a Debugger is available as an item on the menu bar.

13 Miscellaneous
13.1 Stopping
Use CTRL-C to interrupt a command. Use quit to exit from matlab.

13.2 Program-generated names and commands

This shows how to generate variable names ’on the fly’. This is especially useful
when you’re running batch jobs and want to save results into several files.
12 http://www.mathworks.com/access/helpdesk/help/techdoc/matlab external/

12
% A1, A2, A3, and A4 are created and set to 3
for i=1:4
s=sprintf(’A%d= 3’,i);
eval(s);
end;

% This shows how to run a command that’s stored in a string

str=input(’Type in a command ’,’s’)
eval(str)

13.3 Vectorising
for loops are slow and can often be eliminated, though the resulting code may
be less readable. For example, to replace all values < −1 by −1 and all values
> 1 by 1 in a matrix A you can just do A = min(max(M, -1),1). See the Matlab
vectorisation tricks13 page for details.

13.4 Calling Fortran or C

To find out how to run Fortran or C code from matlab see the External Interfaces14
page which includes tutorials and examples.

13
http://www-h.eng.cam.ac.uk/help/tpl/programs/Matlab/tricks.html
14 http://www.mathworks.com/access/helpdesk/help/techdoc/matlab external/matlab external.shtml

13