Matlab Brief Tutorial: Introductory Concepts

Matlab brief tutorial
Table of Contents
Introductory concepts ........................................................................................................... 1
Matrix generation ................................................................................................................ 4
Matrix operations ................................................................................................................ 9
Array indexing .................................................................................................................. 14
Workspace ........................................................................................................................ 22
Strings ............................................................................................................................. 22
Functions ......................................................................................................................... 23
Plots ................................................................................................................................ 24
Scripting .......................................................................................................................... 37
Loops and conditional statements ......................................................................................... 39
Final hint ......................................................................................................................... 45
Introductory Matlab script by Giulio Coluccia (giulio.coluccia@polito.it) Main source: Matlab Quick Start Guide by
Mathworks http://www.mathworks.it/help/techdoc/learn_matlab/bta3so7.html
An up-to-date primer to Matlab is available here: http://www.mathworks.it/help/pdf_doc/matlab/getstart.pdf
Introductory concepts
Matlab is a suite for Numerical Analysis . Its capabilities can be expanded by using Toolboxes (e.g. for
Symbolic Analysis). This means that each variable must be assigned a value:
a = 5 % Generate a variable called 'a' and assign it a value of 5
a % Ask for a's value
a =
a =
Matlab is an interpreted language. This means that there is no need to compile. Errors are discovered at
runtime. E.g. let's ask the value of variable b
This generates an error
% b
Let's define it
b = 3
1
b =
Usual operators have usual purposes...
c = a+b
d = a/b
c =
d =
1.6667
as usual mathematatical functions
e = sin(c)
f = sin(d)
e =
0.9894
f =
0.9954
if no output variable is specified, Malab creates the variable ans for you
a*b
ans =
15
To suppress the output of a command on screen, use ;
e = a-b;
i and j denote complex unit = sqrt(-1). It's better not to use them for naming variables.
z = a+j*b
y = c+i*d
2
real(z), imag(y), abs(z), angle(y)
z =
5.0000 + 3.0000i
y =
8.0000 + 1.6667i
ans =
ans =
1.6667
ans =
5.8310
ans =
0.2054
Matlab knows a certain number of constants
pi
eps
Inf
NaN
ans =
3.1416
ans =
2.2204e-16
ans =
Inf
3
ans =
NaN
Variables can have any alphanumerical name starting with a letter. Matlab is case sensitive
abc123 = 6;
This exists
abc123
abc123 =
This one doesn't and generates an error
% Abc123
Up and down arrows can be used to recall previously issued commands. If you press them after typing,
only commands in history beginning with what you typed will be shown.
Matrix generation
Matlab is optimized for matrix calculations. Actually, scalars are treated as 1x1 matrices, column vectors
as Nx1 matrices and row vectors as 1xN matrices.
Matlab can also work with multi-dimensional matrices, i.e. matrices with D>2 dimensions.
Matrices can be created:
1. Manually inserting its elements
2. From built-in functions
3. From M-files (later)
4. Loaded from external data files (later)
Let's create a 3x3 matrix. Matrices are enclosed by square brackets. Space or , separate elements in a row.
; separates rows from each other
A = [1 2 3; 4 5 6; 7, 8, 9]
A =
1 2 3
4 5 6
7 8 9
4
Row vector
B = [4 5 6]
B =
4 5 6
Column vector
C = [7;8;9]
C =
7
8
9
Let's create a 2x3 matrix of zeros
zeri = zeros(2,3)
zeri =
0 0 0
0 0 0
What is |zeros|? Let's ask for help
help zeros
ZEROS Zeros array.

ZEROS(N) is an N-by-N matrix of zeros.
ZEROS(M,N) or ZEROS([M,N]) is an M-by-N matrix of zeros.
ZEROS(M,N,P,...) or ZEROS([M N P ...]) is an M-by-N-by-P-by-... array o

zeros.
ZEROS(SIZE(A)) is the same size as A and all zeros.
ZEROS with no arguments is the scalar 0.
ZEROS(..., CLASSNAME) is an array of zeros of class specified by the

string CLASSNAME.
ZEROS(..., 'like', Y) is an array of zeros with the same data type, spa
and complexity (real or complex) as the numeric variable Y.
Note: The size inputs M, N, and P... should be nonnegative integers.
5
Negative integers are treated as 0.
Example:
x = zeros(2,3,'int8');
See also EYE, ONES.
Overloaded methods:
distributed/zeros
codistributor2dbc/zeros
codistributor1d/zeros
codistributed/zeros
gpuArray/zeros
Reference page in Help browser

doc zeros
Let's create a 4x4 matrix of uniformly distributed elements
unif1 = rand(4,4)
unif1 =
0.4995 0.3383 0.3761 0.9319

0.6434 0.5756 0.2587 0.0235
0.8723 0.8703 0.7623 0.6562
0.1314 0.1272 0.5571 0.6566
but also
unif2 = rand(4)
unif2 =
0.8729 0.2509 0.7414 0.5307

0.9315 0.2515 0.2400 0.7222
0.8860 0.6849 0.4127 0.0580
0.4591 0.2608 0.8285 0.0926
because matlab functions can have different syntaxes
help rand
RAND Uniformly distributed pseudorandom numbers.

R = RAND(N) returns an N-by-N matrix containing pseudorandom values dra
from the standard uniform distribution on the open interval(0,1). RAND
or RAND([M,N]) returns an M-by-N matrix. RAND(M,N,P,...) or
RAND([M,N,P,...]) returns an M-by-N-by-P-by-... array. RAND returns a
scalar. RAND(SIZE(A)) returns an array the same size as A.
Note: The size inputs M, N, P, ... should be nonnegative integers.
6
Negative integers are treated as 0.
R = RAND(..., CLASSNAME) returns an array of uniform values of the

specified class. CLASSNAME can be 'double' or 'single'.
R = RAND(..., 'like', Y) returns an array of uniform values of the

same class as Y.
The sequence of numbers produced by RAND is determined by the settings

the uniform random number generator that underlies RAND, RANDI, and RAN
Control that shared random number generator using RNG.
Examples:
Example 1: Generate values from the uniform distribution on the

interval [a, b].
r = a + (b-a).*rand(100,1);
Example 2: Use the RANDI function, instead of RAND, to generate

integer values from the uniform distribution on the set 1:100.
r = randi(100,1,5);
Example 3: Reset the random number generator used by RAND, RANDI, an

RANDN to its default startup settings, so that RAND produces the sam
random numbers as if you restarted MATLAB.
rng('default')
rand(1,5)
Example 4: Save the settings for the random number generator used by
RAND, RANDI, and RANDN, generate 5 values from RAND, restore the
settings, and repeat those values.
s = rng
u1 = rand(1,5)
rng(s);
u2 = rand(1,5) % contains exactly the same values as u1
Example 5: Reinitialize the random number generator used by RAND,

RANDI, and RANDN with a seed based on the current time. RAND will
return different values each time you do this. NOTE: It is usually
not necessary to do this more than once per MATLAB session.
rng('shuffle');
rand(1,5)
See <a href="matlab:helpview([docroot '\techdoc\math\math.map'],'update

RAND with the 'seed', 'state', or 'twister' inputs.
See also RANDI, RANDN, RNG, RANDSTREAM, RANDSTREAM/RAND,

SPRAND, SPRANDN, RANDPERM.
Overloaded methods:
RandStream/rand
distributed/rand
codistributor2dbc/rand
codistributor1d/rand
7
codistributed/rand
gpuArray/rand
Reference page in Help browser

doc rand
other useful matrix building functions: randn, eye, ones
generate the empty matrix
Aempty = []
Aempty =
[]
generate "range" vectors with : 1 to 20
a1to20 = 1:20
a1to20 =
Columns 1 through 13
1 2 3 4 5 6 7 8 9 10 11 12
14 15 16 17 18 19 20
1 to 20 with step size of 2
a1to20step2 = 1:2:20
a1to20step2 =
1 3 5 7 9 11 13 15 17 19
steps can be negative
a20to1stepNeg = 20:-1:1
a20to1stepNeg =
20 19 18 17 16 15 14 13 12 11 10 9
8
7 6 5 4 3 2 1
Matrix operations
Matlab allows to process all the elements of a matrix as a whole
a = randn(3,4)
a + 5
sin(a)
a =
-0.8380 -0.1900 -0.0787 -1.7703

-1.1438 0.4207 -1.3214 0.7532
-2.2245 0.7608 0.3525 -0.8529
ans =
4.1620 4.8100 4.9213 3.2297

3.8562 5.4207 3.6786 5.7532
2.7755 5.7608 5.3525 4.1471
ans =
-0.7433 -0.1888 -0.0786 -0.9802

-0.9102 0.4084 -0.9691 0.6839
-0.7938 0.6895 0.3453 -0.7532
+ and - perform matrix sum and subtraction
b = randn(3,4)
c = a+b
d = a-b
b =
0.9724 -1.0063 -1.3667 1.4398

-1.0553 -1.1227 0.7456 0.6397
-1.0872 -1.5828 -0.9520 -1.0355
c =
0.1344 -1.1962 -1.4454 -0.3306

-2.1992 -0.7020 -0.5759 1.3929
-3.3118 -0.8219 -0.5995 -1.8884
9
d =
-1.8104 0.8163 1.2881 -3.2101

-0.0885 1.5434 -2.0670 0.1134
-1.1373 2.3436 1.3045 0.1825
dimensions must be compliant
e = ones(4,2);
This generates an error
% f = a+e
* performs matrix product (row-column product)
f = b*e
f =
0.0392 0.0392
-0.7928 -0.7928
-4.6575 -4.6575
dimensions must be compliant. This generates an error
% g = a*b
\ perform "matrix left division", which is the solution to Ax = B, with A square and invertible and x =
A(^-1)*B = A\B
A = randn(5)
B = randn(5,1)
x1 = A\B
x2 = inv(A)*B % inv() performs matrix inversion
A =
0.8774 -0.1936 0.5712 0.6825 -1.9796

-1.1566 -0.1576 0.5184 -0.7971 -0.6150
-0.4346 0.9342 0.2793 -1.8054 -0.5038
0.1135 -1.7825 -1.3907 -0.0204 1.3604
0.9226 0.9238 0.4709 -1.0997 -0.3535
B =
-1.6036
0.8091
10
1.7550
-0.0805
2.6687
x1 =
0.9623
-0.8440
2.7553
-1.6458
1.5467
x2 =
0.9623
-0.8440
2.7553
-1.6458
1.5467
|'| takes the hermitian of a matrix. |.'| transposes it.
z = a+j*b % j is a constant
z_herm = z'
z_trasp = z.'
z =
-0.8380 + 0.9724i -0.1900 - 1.0063i -0.0787 - 1.3667i -1.7703 + 1.4398

-1.1438 - 1.0553i 0.4207 - 1.1227i -1.3214 + 0.7456i 0.7532 + 0.6397
-2.2245 - 1.0872i 0.7608 - 1.5828i 0.3525 - 0.9520i -0.8529 - 1.0355
z_herm =
-0.8380 - 0.9724i -1.1438 + 1.0553i -2.2245 + 1.0872i

-0.1900 + 1.0063i 0.4207 + 1.1227i 0.7608 + 1.5828i
-0.0787 + 1.3667i -1.3214 - 0.7456i 0.3525 + 0.9520i
-1.7703 - 1.4398i 0.7532 - 0.6397i -0.8529 + 1.0355i
z_trasp =
-0.8380 + 0.9724i -1.1438 - 1.0553i -2.2245 - 1.0872i

-0.1900 - 1.0063i 0.4207 - 1.1227i 0.7608 - 1.5828i
-0.0787 - 1.3667i -1.3214 + 0.7456i 0.3525 - 0.9520i
-1.7703 + 1.4398i 0.7532 + 0.6397i -0.8529 - 1.0355i
Array operators perform elementwise operations
11
Elementwise product
ab = a.*b
ab =
-0.8149 0.1912 0.1075 -2.5488

1.2072 -0.4723 -0.9852 0.4818
2.4186 -1.2042 -0.3356 0.8832
Elementwise division
ba = a./b
ba =
-0.8617 0.1888 0.0576 -1.2296

1.0839 -0.3747 -1.7724 1.1773
2.0460 -0.4807 -0.3703 0.8237
Elementwise power
a5 = A.^5
a5 =
0.5201 -0.0003 0.0608 0.1481 -30.4035

-2.0701 -0.0001 0.0374 -0.3219 -0.0880
-0.0155 0.7117 0.0017 -19.1827 -0.0324
0.0000 -17.9931 -5.2024 -0.0000 4.6599
0.6684 0.6726 0.0232 -1.6082 -0.0055
which is different from ^
ampow5 = A^5
ampow5 =
26.0089 -15.7019 -2.6425 20.2736 -18.7352

-24.2961 16.4000 12.3783 -19.8470 -6.2211
-23.5512 29.7664 20.7310 -35.3572 -7.4780
19.0808 -30.2084 -22.5504 22.6103 15.7587
-0.0258 19.1475 11.5044 -19.3778 -1.2395
Same result on real matrices
a_herm = a'
12
a_trasp = a.'
a_herm =
-0.8380 -1.1438 -2.2245

-0.1900 0.4207 0.7608
-0.0787 -1.3214 0.3525
-1.7703 0.7532 -0.8529
a_trasp =
-0.8380 -1.1438 -2.2245

-0.1900 0.4207 0.7608
-0.0787 -1.3214 0.3525
-1.7703 0.7532 -0.8529
size of a matrix
size(a)
size(a,1) % Rows
size(a,2) % Columns
size(Aempty)
ans =
3 4
ans =
ans =
ans =
0 0
Matrices can be concatenated like elements of a matrix
Conc = [randn(4) ones(4,2); eye(3) zeros(3)]
Conc =
-0.3198 -0.4092 -0.1084 -0.6874 1.0000 1.0000
13
-0.5632 1.2922 0.6298 0.0881 1.0000 1.0000

0.7212 0.7626 1.0398 -0.7203 1.0000 1.0000
-0.2140 3.5214 0.2222 0.3207 1.0000 1.0000
1.0000 0 0 0 0 0
0 1.0000 0 0 0 0
0 0 1.0000 0 0 0
Which size will Conc have?
size(Conc)
ans =
7 6
Array indexing
Read the element of A positioned on 5th row and 8th column
A = randn(5,8)
A =
Columns 1 through 7
2.1277 0.3068 -1.4784 0.6193 1.1220 -0.1772 -0.7893

-2.3367 -0.7322 0.5752 0.4263 0.9210 -2.6973 0.4241
-0.0381 0.4215 -0.9627 0.1025 0.4017 0.3031 0.5714
-0.5468 0.8896 0.6207 -0.9929 0.9076 -1.4990 0.4055
-0.9498 -0.7648 0.0968 -0.8502 -1.3678 0.4723 1.5286
Column 8
0.6117
0.3538
0.7736
1.5026
0.3575
indexes start from 1 (not like C, C++ ...)
address a single element with brackets. Read
a34 = A(3,4)
a34 =
0.1025
14
it must be within matrix size, otherwise an error is generated
% a39 = A(3,9)
address a single element with brackets. Assign
A(4,5) = 100
A =
Columns 1 through 7
2.1277 0.3068 -1.4784 0.6193 1.1220 -0.1772 -0.7893

-2.3367 -0.7322 0.5752 0.4263 0.9210 -2.6973 0.4241
-0.0381 0.4215 -0.9627 0.1025 0.4017 0.3031 0.5714
-0.5468 0.8896 0.6207 -0.9929 100.0000 -1.4990 0.4055
-0.9498 -0.7648 0.0968 -0.8502 -1.3678 0.4723 1.5286
Column 8
0.6117
0.3538
0.7736
1.5026
0.3575
it can lie outside matrix size. Matrix will be padded with 0s
A(3,9) = 555
A =
Columns 1 through 7
2.1277 0.3068 -1.4784 0.6193 1.1220 -0.1772 -0.7893

-2.3367 -0.7322 0.5752 0.4263 0.9210 -2.6973 0.4241
-0.0381 0.4215 -0.9627 0.1025 0.4017 0.3031 0.5714
-0.5468 0.8896 0.6207 -0.9929 100.0000 -1.4990 0.4055
-0.9498 -0.7648 0.0968 -0.8502 -1.3678 0.4723 1.5286
Columns 8 through 9
0.6117 0
0.3538 0
0.7736 555.0000
1.5026 0
0.3575 0
Address a row. : means "every"
15
A(3,:)
ans =
Columns 1 through 7
-0.0381 0.4215 -0.9627 0.1025 0.4017 0.3031 0.5714
Columns 8 through 9
0.7736 555.0000
Address a column
A(:,4)
ans =
0.6193
0.4263
0.1025
-0.9929
-0.8502
indexes are actually vectors
A(2:4,3:5)
ans =
0.5752 0.4263 0.9210

-0.9627 0.1025 0.4017
0.6207 -0.9929 100.0000
all even rows. Note the use of keyword end wich means "last element of"
A(2:2:end,:)
ans =
Columns 1 through 7
-2.3367 -0.7322 0.5752 0.4263 0.9210 -2.6973 0.4241

-0.5468 0.8896 0.6207 -0.9929 100.0000 -1.4990 0.4055
Columns 8 through 9
0.3538 0
16
1.5026 0
take the 3rd, the 6th and the 7th element of 5th row
A(5,[3 6 7])
ans =
0.0968 0.4723 1.5286
vectorize a matrix
A_vec = A(:)
size(A_vec)
A_vec =
2.1277
-2.3367
-0.0381
-0.5468
-0.9498
0.3068
-0.7322
0.4215
0.8896
-0.7648
-1.4784
0.5752
-0.9627
0.6207
0.0968
0.6193
0.4263
0.1025
-0.9929
-0.8502
1.1220
0.9210
0.4017
100.0000
-1.3678
-0.1772
-2.6973
0.3031
-1.4990
0.4723
-0.7893
0.4241
0.5714
0.4055
17
1.5286
0.6117
0.3538
0.7736
1.5026
0.3575
0
0
555.0000
0
0
ans =
45 1
vectors are addressed in the same way by a single index
A_vec(5)
A_vec(1:2:end)
ans =
-0.9498
ans =
2.1277
-0.0381
-0.9498
-0.7322
0.8896
-1.4784
-0.9627
0.0968
0.4263
-0.9929
1.1220
0.4017
-1.3678
-2.6973
-1.4990
-0.7893
0.5714
1.5286
0.3538
1.5026
0
555.0000
0
18
Conditional indexing
G = randn(5)
G>0
H = zeros(size(G))
H(G>0) = 10
H(G<0) = -5
%
Pos_Index = find(G>0)
Neg_Index = find(G<0)
G(Pos_Index)
G(G>0)
G(Neg_Index)
G(G<0)
G =
-0.2324 0.3793 0.1371 -0.3835 -0.7224

-0.0883 0.3244 -1.2577 2.2697 2.5006
0.2226 0.4482 -1.0924 0.6960 -0.5684
1.9802 0.1283 -1.1220 -0.1058 0.9096
-0.3559 0.4914 0.7555 0.0726 -0.1725
ans =
0 1 1 0 0
0 1 0 1 1
1 1 0 1 0
1 1 0 0 1
0 1 1 1 0
H =
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
H =
0 10 10 0 0
0 10 0 10 10
10 10 0 10 0
10 10 0 0 10
0 10 10 10 0
19
H =
-5 10 10 -5 -5
-5 10 -5 10 10
10 10 -5 10 -5
10 10 -5 -5 10
-5 10 10 10 -5
Pos_Index =
3
4
6
7
8
9
10
11
15
17
18
20
22
24
Neg_Index =
1
2
5
12
13
14
16
19
21
23
25
ans =
0.2226
1.9802
0.3793
0.3244
0.4482
0.1283
0.4914
0.1371
0.7555
20
2.2697
0.6960
0.0726
2.5006
0.9096
ans =
0.2226
1.9802
0.3793
0.3244
0.4482
0.1283
0.4914
0.1371
0.7555
2.2697
0.6960
0.0726
2.5006
0.9096
ans =
-0.2324
-0.0883
-0.3559
-1.2577
-1.0924
-1.1220
-0.3835
-0.1058
-0.7224
-0.5684
-0.1725
ans =
-0.2324
-0.0883
-0.3559
-1.2577
-1.0924
-1.1220
-0.3835
-0.1058
-0.7224
-0.5684
-0.1725
21
Workspace
All variables are stored in the workspace. Workspace can be saved on file
save myWorkspace
You can clear selected variables
clear A A_vec
or emptied
clear
and restored
load myWorkspace
Strings
Strings must be created within |'|s and can be assigned to variables
myString = 'My string'
myString =
My string
strings are arrays of chars
myString(4)
ans =
so as vectors they can be concatenated
myConcat = [myString ' does not have a meaning']
myConcat =
My string does not have a meaning
respecting sizes: wrong
% ['ciao';'come'; 'va?']
correct
22
['ciao';'come'; 'va? ']
ans =
ciao
come
va?
It's useful to convert numbers to strings with num2str
f = 71;
c = (f-32)/1.8;
tempText = ['Temperature is ',num2str(c),'C']
tempText =
Temperature is 21.6667C
Functions
MATLAB provides a large number of functions that perform computational tasks. Functions are equivalent
to subroutines or methods in other programming languages.
A = [1 3 5];
B = [10 6 4];
To call a function, enclose its input arguments in parentheses:
max(A)
ans =
If there are multiple input arguments, separate them with commas:
max(A,B)
ans =
10 6 5
Return output from a function by assigning it to a variable:
maxA = max(A);
When there are multiple output arguments, enclose them in square brackets
23
[maxA,location] = max(A);
Enclose any character string inputs in single quotes:
disp('hello world'); % disp displays the content of the variable
hello world
To call a function that does not require any inputs and does not return any outputs, type only the function
name:
clc % The clc function clears the Command Window.
Plots
Create a new figure object and the data set. We want to plot sin(x) from 0 to 2*pi in steps of pi/100
figure;
x = 0:pi/50:2*pi;
y_sin = sin(x);
plot(x,y_sin)
We can label the axes
xlabel('x')
ylabel('y')
24
And give the plot a title
title('Plot of the Sine Function')
25
And add a grid overlay
grid on
26
To superimpose different data, we hold the graph and draw the cos in dashed red with diamond marker.
Then we release the plot in the end.
hold on
y_cos = cos(x);
plot(x, y_cos, 'rd--');
hold off
27
We can also add a legend
legend('sin','cos');
28
And focus on a specific axis range
axis([0 pi/2 0 1]);
29
Instead, we may want to generate a 2x1 subplot Generate a 2x1 subplot and place sin plot in slot #1
figure;
subplot(2,1,1);
plot(x,y_sin)
xlabel('x')
ylabel('sin(x)')
title('Plot of the Sine Function')
grid on
30
then move to slot #2 and place cos plot
subplot(2,1,2);
plot(x,y_cos)
xlabel('x')
ylabel('cos(x)')
title('Plot of the Cosine Function')
grid on
31
Other useful plots: histogram
z = randn(10000,1);
figure;
hist(z,100)
32
stem plot
figure;
stem(x, y_sin)
33
logaritmic plot
x = 0:0.01:3;
y = exp(-x.^2); % Note the .^
Linear plot
figure;
plot(x,y);
grid on;
34
Logaritmic on y axis
figure;
semilogy(x,y);
grid on;
35
Logaritmic on both axis
figure;
loglog(x,y)
grid on;
36
close all figures with close all;
Scripting
Scripts are trivially sequences of operations. This file itself is a script. Scripts are ASCII files with .m
extension and can be run typing their filename, if they are placed in current directory or in matlab path
myScript
This is a script
37
Type a script content
type myScript.m
clc;
disp('This is a script');
% This is a comment
% Generate random data from a uniform distribution

% and calculate the mean. Plot the data and the mean.
n = 50; % 50 data points

r = rand(n,1);
plot(r,'ko')
% Draw a line from (0,m) to (n,m)

m = mean(r);
hold on
plot([0,n],[m,m],'--')
hold off
title('Mean of Random Uniform Data')
scripts can be terminated by Ctrl+C
Moreover, custom functions can be written. They have optional inputs and optional outputs.
numberOfData = 10000;
38
a = -5;
b = 9;
[meanData, varianceData] = myFunction(numberOfData, a, b);

disp(['Analytical average is ' num2str((a+b)/2)]);
disp(['Experimental average is ' num2str(meanData)]);
disp(['Analytical variance is ' num2str((b-a).^2./12)]);
disp(['Experimental variance is ' num2str(varianceData)]);
Analytical average is 2
Experimental average is 1.9593
Analytical variance is 16.3333
Experimental variance is 16.5392
You can also provide help to custom functions
help myFunction
type myFunction.m
MYFUNCTION Evaluates average and variance of uniform distribution

numberOfData is the number of samples the evaluation is performed on
minimum is the minimum of uniform distribution
maximum is the minimum of uniform distribution
function [ average, variance ] = myFunction( numberOfData, minimum, maximum

%MYFUNCTION Evaluates average and variance of uniform distribution
% numberOfData is the number of samples the evaluation is performed on
% minimum is the minimum of uniform distribution
% maximum is the minimum of uniform distribution
x = (maximum-minimum)*rand(numberOfData, 1) + minimum;
average = mean(x);
variance = var(x);
end
Loops and conditional statements

Loops are created using for keyword (also while exists). Both are terminated with an end
a = -5;
b = 9;
for numberOfData = 1:100:1000
[meanData, varianceData] = myFunction(numberOfData, a, b)
end
meanData =
39
2.6559
varianceData =
meanData =
1.8471
varianceData =
13.5024
meanData =
2.0073
varianceData =
15.5088
meanData =
2.1701
varianceData =
17.0583
meanData =
1.9149
varianceData =
15.7605
meanData =
2.4773
varianceData =
40
16.4590
meanData =
1.9687
varianceData =
15.4077
meanData =
1.9147
varianceData =
16.4750
meanData =
2.1541
varianceData =
15.8925
meanData =
1.9696
varianceData =
15.6236
adding some output

disp('=================================================================');
disp(['number of data is ' num2str(numberOfData)]);
disp(['Analytical variance is ' num2str((b-a).^2./12)]);
disp(['Experimental variance is ' num2str(varianceData)]);
41
end
=================================================================
number of data is 1
Experimental variance is 0
=================================================================
number of data is 101
=================================================================
=================================================================
=================================================================
=================================================================
=================================================================
=================================================================
=================================================================
42

=================================================================
and an |if| statement

if(condition)
if true
else
if false
end
% Relation operators are ==, ~=, >, <, >=, <=

% Logical operators are &, |, ~

disp('=================================================================');
disp(['number of data is ' num2str(numberOfData)]);
if(meanData > (a+b)/2)
disp('Experimental above analytical');
elseif(meanData == (a+b)/2)
disp('Experimental exactly equal to analytical');
else
disp('Experimental below analytical');
end
end
=================================================================
number of data is 1
Experimental average is -4.1981
Experimental below analytical
=================================================================
Experimental above analytical
=================================================================
=================================================================
43
=================================================================
=================================================================
=================================================================
=================================================================
=================================================================
=================================================================
Final example
numberOfData = 1:100:1000;
meanData = zeros(size(numberOfData)); % optional
for ii = 1:length(numberOfData)
[meanData(ii), varianceData] = myFunction(numberOfData(ii), a, b);
end
figure;
plot(numberOfData, meanData,'rx');
hold on;
plot([numberOfData(1) numberOfData(end)], [(a+b)/2 (a+b)/2]);
44
Final hint
Matlab works very well with matrix/vector calculations. When possible, use matrix operations rather than
for loops
A is a NxN square invertible matrix X is an unknown NxM matrix B is a given NxM matrix Find the 1xM
vector of column-averages of X such that X is the solution to A*X = B
N = 100;
M = 10000;
A = randn(N);
B = randn(N,M);
Bad Solution:
tic;
for ii = 1:M
b_temp = B(:,ii);
x_temp = inv(A)*b_temp;
averages_bad(ii) = mean(x_temp);
end
toc
Elapsed time is 4.260555 seconds.
Intermediate solution
45
tic;
averages_interm = zeros(1,M);
inv_A = inv(A);
for ii = 1:M
b_temp = B(:,ii);
x_temp = inv_A*b_temp;
averages_interm(ii) = mean(x_temp);
end
toc
Best solution
tic;
averages_best = mean(A\B);
toc
% disp([averages_bad.' averages_interm.' averages_best.']);
Published with MATLAB® R2014a
46

Matlab Brief Tutorial: Introductory Concepts

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Matlab Brief Tutorial: Introductory Concepts

Transféré par

Droits d'auteur :

Formats disponibles

Matlab brief tutorial

An up-to-date primer to Matlab is available here: http://www.mathworks.it/help/pdf_doc/matlab/getstart.pdf

a = 5 % Generate a variable called 'a' and assign it a value of 5

a % Ask for a's value

This generates an error

Usual operators have usual purposes...

as usual mathematatical functions

To suppress the output of a command on screen, use ;

real(z), imag(y), abs(z), angle(y)

Matlab knows a certain number of constants

This one doesn't and generates an error

Matrices can be created:

1. Manually inserting its elements

2. From built-in functions

3. From M-files (later)

4. Loaded from external data files (later)

Let's create a 2x3 matrix of zeros

What is |zeros|? Let's ask for help

ZEROS Zeros array.

ZEROS(M,N) or ZEROS([M,N]) is an M-by-N matrix of zeros.

ZEROS(M,N,P,...) or ZEROS([M N P ...]) is an M-by-N-by-P-by-... array o

ZEROS(SIZE(A)) is the same size as A and all zeros.

ZEROS with no arguments is the scalar 0.

ZEROS(..., CLASSNAME) is an array of zeros of class specified by the

Note: The size inputs M, N, and P... should be nonnegative integers.

Negative integers are treated as 0.

See also EYE, ONES.

Reference page in Help browser

Let's create a 4x4 matrix of uniformly distributed elements

0.4995 0.3383 0.3761 0.9319

0.8729 0.2509 0.7414 0.5307

because matlab functions can have different syntaxes

RAND Uniformly distributed pseudorandom numbers.

Note: The size inputs M, N, P, ... should be nonnegative integers.

Negative integers are treated as 0.

R = RAND(..., CLASSNAME) returns an array of uniform values of the

R = RAND(..., 'like', Y) returns an array of uniform values of the

The sequence of numbers produced by RAND is determined by the settings

Example 1: Generate values from the uniform distribution on the

Example 2: Use the RANDI function, instead of RAND, to generate

Example 3: Reset the random number generator used by RAND, RANDI, an

Example 5: Reinitialize the random number generator used by RAND,

See <a href="matlab:helpview([docroot '\techdoc\math\math.map'],'update

See also RANDI, RANDN, RNG, RANDSTREAM, RANDSTREAM/RAND,

Reference page in Help browser

other useful matrix building functions: randn, eye, ones

generate the empty matrix

generate "range" vectors with : 1 to 20

1 to 20 with step size of 2

steps can be negative

-0.8380 -0.1900 -0.0787 -1.7703

4.1620 4.8100 4.9213 3.2297

-0.7433 -0.1888 -0.0786 -0.9802

+ and - perform matrix sum and subtraction

0.9724 -1.0063 -1.3667 1.4398

0.1344 -1.1962 -1.4454 -0.3306

-1.8104 0.8163 1.2881 -3.2101

dimensions must be compliant

This generates an error

* performs matrix product (row-column product)

dimensions must be compliant. This generates an error

0.8774 -0.1936 0.5712 0.6825 -1.9796

|'| takes the hermitian of a matrix. |.'| transposes it.