MATLAB: An introduction

with applications
3. Mathematical
Operations with Arrays
 Context of This Chapter
• Addition and subtraction of arrays

• Multiplication, division, and exponentiation

• Element-by-element operations
 Addition and Subtraction
 a11 a12 a13  b11 b12 b13   a11  b11 a12  b12 a13  b13 
A= , B=   AB =  
 21 22 23 
a a a  21 22 23 
b b b a  b a 
 21 21 22 22 23 23 
b a  b

>> v = [3 6 2]; w=[9 -1 5]; >> D = A + B

>> z = v + w D = 12 4 11 When a scalar is
z= -3 19 6 added to an array,
12 5 7 >> D-2 the scalar is
>> A=[2 -3 7; 8 4 5]; ans = 10 2 9 added to all the
>> B=[10 7 4; -11 15 1]; -5 17 4 elements of the
>> C = A - B >> C * 2 array.
C= ans =
-8 -10 3 -16 -20 6
19 -11 4 38 -22 8
 Array Multiplication
• The multiplication operation * is executed by MATLAB according
to the rules of linear algebra.
b11 b12 
 a11 a12 a13   
A=  , B = b21 b22 
a21 a22 a23 
b31
 b32 
 a11b11 + a12b21 + a13b31 a11b12 + a12b22 + a13b32 
 AB =  
a21b11 + a22b21 + a23b31 a21b12 + a22b22 + a23b32 

>> A = [ 1 4 3; 2 6 1; 5 2 8];
>> B = [5 4; 1 3; 2 6]; 1 4 3 5 4 
   
>> C = A*B  2 6 1 
 1 3 
5 2 8 2 6
T= 15 34    
18 32 1 5 + 4 1 + 3  2 1 4 + 4  3 + 3  6  15 34 
43 74    
= 2  5 + 6 1 + 1 2 2  4 + 6  3 + 1 6  = 18 32 
>> D = B*A 5  5 + 2 1 + 8  2 5  4 + 2  3 + 8  6 43 74
??? Error using ==> mtimes    
Inner matrix dimensions must agree.
 Array Multiplication
• The multiplication of matrices is not commutative.
A*B ≠ B*A

 w1   w1   w1v1 w1v2 w1v3 

     
v1 v2 v3   w2  = v1w1 + v2 w2 + v3 w3 ,  w2   v1 v2 v3  =  w2 v1 w2 v2 w2 v3 
 w3   w3   w3v1 w3v3 
     w3v2

>> A = [ 6 2; 4 7]; B = [3 9; 5 1];

>> C = A*B, D = B*A % A*B≠B*A
C=
28 56
47 43
D=
54 69
34 17
 Array Multiplication
>> v = [ 6 2 4 ]; w = [3; 9; 5]; >> G*F
>> x = v*w ans =
14 26
x=
v*w 31 45
56
≠ w*v >> A=[2 5 7 0; 10 1 3 4; 6 2 11 5]
>> z = w*v
A=
z= 2 5 7 0
18 6 12 10 1 3 4
54 18 36 6 2 11 5
30 10 20 >> C = 3*A
>> dot(v, w) % v*w C=
ans = 6 15 21 0
56 30 3 9 12
>> F=[1 3; 5 7]; G=[4 2; 1 6]; 18 6 33 15
>> F*G >> D=A*3
ans = D=
7 20 6 15 21 0
27 52 30 3 9 12
18 6 33 15
A System of Linear Equations
• Linear algebra rules of array multiplication provide a
convenient way for writing a system of linear
equations.

a11 x1 + a12 x2 + a13 x3 = b1  a11 a12 a13   x1   b1 

a21 x1 + a22 x2 + a23 x3 = b2 a a23   x  = b   Ax = b
 21 a22  2  2
a31 x1 + a32 x2 + a33 x3 = b3  a31 a32 a33   x3  b3 
 Array Division
• Identity matrix: eye(n)
– A square matrix in which the diagonal
elements are 1s, and the rest of the
elements are 0s.

– AI=IA=A

1 0 0  1 0 0   7   7 
6 2 9    6 2 9  0 1 0   3  =  3 
5 10 4  0 1 0  = 5 10 4  ,      
      0 0 1  9  9 
 0 0 1 
 Array Division
• Inverse of a matrix
– The matrix B is the inverse of the matrix A if, when the two
matrices are multiplied, the product is the identity matrix.
– BA=AB=I
– A^-1, or inv(A)

>> A= [ 7 4 6; 3 1 8; 2 5 4]; >> A^-1 % inverse of A

>> B=inv(A) ans =
B= 0.2278 -0.0886 -0.1646
0.2278 -0.0886 -0.1646 -0.0253 -0.1013 0.2405
-0.0253 -0.1013 0.2405 -0.0823 0.1709 0.0316
-0.0823 0.1709 0.0316
>> A*B % equals to B*A >> A*A^-1 % AA-1=I
5.55 e-17
ans = ans =
1.0000 0.0000 0.0000 1.0000 0.0000 0.0000
0 1.0000 0 0 1.0000 0
-0.0000 0.0000 1.0000 -0.0000 0.0000 1.0000
 Array Division
• Determinants
– A matrix has an inverse only if it is square and its
determinant is not equal to zero.
– Denoted by det(A) or |A|
a11 a12
A= = a11a22 − a12 a21
a21 a22
a11 a12 a13
B = a21 a22 a23 = a11a22 a33 + a12 a23 a31 + a13 a21a32 − a23 a32 a11 − a33 a21a12
a31 a32 a33
1 2 0
>> A= [ 1 2 0; 0 1 1; 3 0 1]; det(A)  
A = 0 1 1 
ans = 3 0 1 
7  
>> B=[ 1 2 1; 0 1 1; 3 1 1]; det(B) 1 2 1
 
ans = B = 0 1 1
3 3 1 1
 
 Array Division
• Left division, \
– AX=B, A-1AX=A-1B
– X=A-1B
– X=A\B

– The solution X is obtained numerically with a method

based on Gauss elimination.
– The left division is recommended for solving a set of
linear equations because the calculation of the inverse
may be less accurate than the Gauss elimination method
when large matrices are involved.
 Array Division
• Right division, /
– XC=D, X and D are row vectors.
– X•CC-1=D•C-1
– X=D•C-1

– In MATLAB the last equation can be written using

the right division character:
– X=D/C
 Array Division
• Use matrix operations to solve the following system of linear equations.
4x – 2y + 6z = 8
2x + 8y + 2z = 4
6x + 10y + 3z = 0
The above system of equations can be written in the matrix form
AX=B or in the form XC=D
 4 −2 6   x  8   4 2 6
 2 8 2  y  =  4 z   −2 8 10 = 8 4 0
      or x y
6 10 3  z   0   6 2 3 

>> A=[ 4 -2 6; 2 8 2; 6 10 3]; b=[8; 4; 0]; Ax = b

>> x=A\ b % left division → A-1Ax=A-1b
x = -1.8049 → x=A\b
0.2927
xC=w
2.6341
→ xCC-1=wC-1
>> C=[4 2 6;-2 8 10; 6 2 3]; w = [8 4 0];
→ x = w/C
>> x = w/C % right division
x = -1.8049 0.2927 2.6341
 Exercises
• Create the following three matrices:

3 5 − 2  − 1 4 1 1 2 6 
A = 4 2 − 6 B = − 2 5 6 C = 3 0 − 3
1 0 5   0 7 8 3 5 − 2

• Use the matrices A,B, and C to answer the following:

• (a) Does A*B=B*A?
• (b) Does (A*B)t=Bt*At? (t means transpose)
• (c) Does (A+B)t=At+Bt
 Exercises
• Solve the following system of three linear equations:

4x-2.5y+3z=22.2
6x-4y-2z=16.6
2x+4y-5.5z=-11.3
 Exercises
• Solve the following system of three linear equations:

4x-2.5y+3z=22.2
6x-4y-2z=16.6
2x+4y-5.5z=-11.3

A=[4 -2.5 3; 6 -4 -2; 2 4 -5.5]

b=[22.2; 16.6; -11.3]
x=A\b or inv(A)*b
Or
C=[4 6 2;-2.5 -4 4;3 -2 -5.5],D=[22.2,16.6,-11.3]
x=D/C
 Exercises
• Solve the following system of five linear equations:

-u+0.5v+2w-1.75x+y=9.5
3u+12v-9w+1.5x-6y=-4.5
3u-1.5v+w+1.25x+0.5y=-11.5
3u+2v-w+1.5x-3y=-3.5
6u+3v+2w+x+8y=-23.5
 Exercises
• Solve the following system of five linear equations:

-u+0.5v+2w-1.75x+y=9.5
3u+12v-9w+1.5x-6y=-4.5
3u-1.5v+w+1.25x+0.5y=-11.5
3u+2v-w+1.5x-3y=-3.5
6u+3v+2w+x+8y=-23.5

A=[-1 0.5 2 -1.75 1; 3 12 -9 1.5 -6; 3 -1.5 1 1.25 0.5;3 2 –1 1.5

-3; 6 3 2 1 8]
b=[9.5; -4.5; -11.5; -3.5; -23.5]
x=A\b or inv(A)*b
 Element-by-Element Operations
• Element-by-element operations can be done only with arrays of
the same size.

• .* (Multiplication), .^(Exponentiation), ./(Right division), .\(Left

division)

• If two vectors v and w are v=[v1, v2, v3], w=[w1 w2 w3]

v.*w = [ v1w1, v2w2, v3w3 ] v./w = [ v1/w1, v2/w2, v3/w3 ]
v.^w = [ v1w1, v2w2, v3w3 ]
For two matrices A and B,
 a11 a12 a13   b11 b12 b13   a11b11 a12b12 a13b13 
A.* B =  a21 a22 a23  b21 b22 b23  =  a21b21 a22b22 a23b23 
 a31 a32 a33  b31 b32 b33   a31b31 a32b32 a33b33 
 Element-by-Element Operations
 a11 a12 a13  b11 b12 b13   a11 / b11 a12 / b12 a13 / b13 
     
A. / B = a21 a22 a23  ./ b21 b22 b23  = a21 / b21 a22 / b22 a23 / b23 
a a33  b31 b32 b33   a31 / b31 a33 / b33 
 31 a32 a32 / b32
 a11n a12n a13n 
 n n 
A.^ n = a21 n
a22 a23  >> C=A./B
a n n
a32 n 
a33 C=
 31  2.0000 1.5000 0.3000
1.6667 4.0000 0.5714
 2 6 3 >> A=[2 6 3; 5 8 4];
A=  >> 2.^B
 5 8 4  >> B=[1 4 10; 3 2 7];
ans =
>> A.*B
1 4 10 ans = 2 16 1024
B= 
2 24 30 8 4 128
3 2 7 
15 16 28 >> A*B
??? Error using==> mtimes
Inner matrix dimensions must agree.
 Element-by-Element Operations
• Element-by-element calculations are very useful for
calculating the value of a function at many values of
its argument.
• y(x)= 2x2- 5x+1
>> x = 1:7
x=
1 2 3 4 5 6 7
>> y = 2*x.^2 – 5*x + 1
y=
-2 -1 4 13 26 43 64

• y(x) = (x3 + 5x)/(4x2 – 10)

>> x = [1:0.03:3];
>> y = (x.^3 + 5*x)./(4*x.^2 - 10);
 Element-by-Element Operations
• Element-by-element calculations are very useful for
calculating the value of a function at many values of
its argument. y=2x - 5x + 1
70
2

• y(x)= 2x2- 5x+1 60

50

>> x = 1:7 40
30

y
x=
20
1 2 3 4 5 6 7 10
>> y = 2*x.^2 – 5*x + 1 0

y= -10
2 4 6
x
-2 -1 4 13 26 43 64
y=(x.3 + 5x)/(4x2 - 10)
60
>> plot(x, y) 40

• y(x) = (x3 + 5x)/(4x2 – 10) 20

-20
y
>> x = [1:0.03:3]; -40

-60
>> y = (x.^3 + 5*x)./(4*x.^2 - 10); -80

>> plot(x, y) -100

1 1.5 2 2.5 3
x
Using Arrays in MATLAB Built-In
Math Functions
• When the argument is an array, the operation that is
defined by the function is executed on each element
of the array.
>> x=[0:pi/6:2*pi]
x= 1

0.8

Columns 1 through 7 0.6

0 0.5236 1.0472 1.5708 2.0944 2.6180 3.1416 0.4

0.2

Columns 8 through 13 0

3.6652 4.1888 4.7124 5.2360 5.7596 6.2832 -0.2

-0.4

>> y=cos(x) -0.6

y= -0.8

-1

Columns 1 through 7 0 1 2 3 4 5 6

1.0000 0.8660 0.5000 0.0000 -0.5000 -0.8660 -1.0000

Columns 8 through 13
-0.8660 -0.5000 -0.0000 0.5000 0.8660 1.0000
>> plot(x, y), grid on
Using Arrays in MATLAB Built-In
Math Functions
• Function sqrt(x)

>> d=[1 4 9; 16 25 36; 49 64 81]

d=
1 4 9
16 25 36
49 64 81
>> h=sqrt(d)
h=
1 2 3
4 5 6
7 8 9
 Exercises
• For the function y = 0.8x3 − 2.1x 2 + 0.75 x , calculate the
value of y for the following values of x using element-
by-element operations: -2,-1,0,1,2,3,4

( x 2 − 2)( x + 4)
• For the function y = , calculate the value
of y for the following values 2 xof x using element-by-
element operations: 1,2,3,4,5,6,7
 Exercises
• For the function y = 0.8 x3 − 2.1x 2 + 0.75 x, calculate the value
of y for the following values of x using element-by-
element operations: -2,-1,0,1,2,3,4
• x=-2:4;
• y=0.8*x.^3-2.1*x.^2+0.75*x
• y = -16.3000 -3.6500 0 -0.5500 -0.5000
4.9500 20.6000
( x 2 − 2)( x + 4)
y=
• For the function 2 x , calculate the value of y
for the following values of x using element-by-element
operations: 1,2,3,4,5,6,7
• X=1:7;
• y=(X.^2-2).*(X+4)./(2*sqrt(X))
• y = -2.5000 4.2426 14.1451 28.0000 46.2866
69.4022 97.7038
 Exercises
• Define the vector v=[2 4 6 8 10]. Then use the vector in a
mathematical expression to create the following vectors:

1 1 1 1 1
• (a) a=
2 4 6 8 10 
1 1 1 1 1 
• (b) b= 2
2 42 62 82 102 

• (c) c = 1 2 3 4 5

a=1./v
b=1./v.^2
c=v/2
 Exercises
• Define the vector v=[2 4 6 8 10]. Then use the vector in a
mathematical expression to create the following vectors:
1 1 1 1 1 
• (a) a =  2 4 6 8 10 
 
• (b) b =  2
1 1 1 1 1 
2 42 62 82 102 
• (c) c = 1 2 3 4 5

a=1./v
a = 0.5000 0.2500 0.1667 0.1250 0.1000
b=1./v.^2
b = 0.2500 0.0625 0.0278 0.0156 0.0100
c=v/2
 Exercises
• Define x and y as the vector x=[1 3 5 7 9] and y=[2 5
8 11 14]. Then use them in the following expressions
to calculate z using element-by-element calculations.

xy 2
• (a) z=
x+ y
• z=x.*y.^2./(x+y)

• (b) z = x( x 2 − y ) − ( x − y ) 2
z=x.*(x.^2-2)-(x-y).^2
 Exercises
• Define x and y as the vector x=[1 3 5 7 9] and y=[2 5
8 11 14]. Then use them in the following expressions
to calculate z using element-by-element calculations.

xy 2
z=
• (a) x+ y
• z=x.*y.^2./(x+y)
• z = 1.3333 9.3750 24.6154 47.0556 76.6957
(b) z = x( x − y ) − ( x − y )
2 2
•
• z=x.*(x.^2-2)-(x-y).^2
• z = -2 17 106 313 686
 Exercises
sin x
• Show that lim
x →0
= 1.
x

Do this by first creating a vector that has the

elements 1.5, 1.0, 0.5, 0.01, 0.005, and 0.00005. Then
create a new vector y in which each element is
determined from the elements of x by sin x .
x
Compare the elements of y with the value 1 (use
format long to display the numbers).
 Exercises
sin x
• Show that lim = 1.
x →0 x
Do this by first creating a vector that has the elements 1.5, 1.0, 0.5, 0.01, 0.005, and
0.00005. Then create a new vector y in which each element is determined from the
elements of x by sin x .
x
Compare the elements of y with the value 1 (use format long to display the numbers).

>> x=[1.5 1.0 0.5 0.01 0.005 0.00005];

>> format long
>> y=sin(x)./x
y=

Columns 1 through 4

0.664996657736036 0.841470984807897 0.958851077208406 0.999983333416666

Columns 5 through 6

0.999995833338542 0.999999999583333
• The median is the middle value when the data are arranged in as
cending order(1...........n)

• The mode of a sample is the most frequent value

• Mean: arithmetic mean

• Standard deviation
 Built-In Functions for Analyzing
Arrays
Function Description Example
mean(A) If A is a vector, returns the mean value of the >> A=[5,9,2,4];
elements of the vector >> mean(A)
ans = 5
C=max(A) If A is a vector, C is the largest element in A. If A >> A=[5,9,2,4,11,6,11,1];
is a matrix, C is a row vector containing the >> C=max(A)
C = 11
largest element of each column of A.
[d,n]=max(x) If A is a vector, d is the largest element in A, and >>[d,n]=max(A)
n is the position of the element (the first if d = 11
n= 5
several have the max value.)
min(A) The same as max(A), but for the smallest >> A=[5,9,2,4];
element. >> min(A)
ans = 2
[d,n]=min(A) The same as [d, n]=max(A), but for the smallest
element.
sum(A) If A is a vector, returns the sum of the elements >> A=[5,9,2,4];
of the vector. >> sum(A)
ans = 20
Function Description Example
sort(A) If A is a vector, arranges the elements of the vector in >> A=[5,9,2,4];
ascending order. >> sort(A)
ans = 2 4 5 9
median(A) If A is a vector, returns the median value of the >> median(A)
elements of the vector. ans = 4.5

std(x) If A is a vector, returns the standard deviation of the >> std(A)

elements of the vector. ans = 2.9439

det(A) Returns the determinant of a square matrix A. >> A=[2, 4; 3, 5];

>> det(A)
ans = -2
dot(a,b) Calculates the scalar (dot) product of two vectors a >> a=[1,2,3]; b=[3,4,5];
and b. The vectors can each be row or column vectors. >> dot(a, b), cross(a, b)
ans = 26
cross(a,b) Calculates the cross product of two vectors a and b ans = -5 3 -2
(aXb). The two vectors must have each three elements.
inv(A) Returns the inverse of a square matrix A. >> A=[2,-2,1;3,2,-1;2,-
3,2];
>> inv(A)
ans = 0.2000 0.2000 0
-1.6000 0.4000 1.0000
-2.6000 0.4000 2.0000
 Generation of Random Numbers
• rand, randn, randi
Command Description Example
rand Generates a single random number between 0 >> rand
and 1. ans = 0.2311

rand(1,n) Generates an n-element row vector of random >> a=rand(1,4)

A=
numbers between 0 and 1.
0.6068 0.4860 0.8913
0.7621
rand(n) Generates an nxn matrix with random numbers >>b=rand(3)
between 0 and 1. b=
0.4565 0.4447 0.9218
0.0185 0.6154 0.7382
0.8214 0.7919 0.1763
rand(m,n) Generates an mxn matrix with random numbers >> c=rand(2,3)
between 0 and 1. c=
0.4057 0.9169 0.8936
0.9355 0.4103 0.0579
randperm(n) Generates a row vector with n elements that are >> c=randperm(8)
random permutation of integers 1 through n. ans =
82743651
 Generation of Random Numbers
• Sometimes there is a need for random numbers that
are distributed in an interval other than (0,1), or for
numbers that are integers only.

• Range (a,b): (b-a)*rand+a

• A vector of 10 elements with random values between

-5 and 10 (a=-5, b=10)
>> v=15*rand(1,10)-5
v = 7.2209 8.5869 -3.0952 8.7006 4.4854 -3.5369 -0.8225 3.2032 9.3626
9.4733
 Randi
• The randi command generates uniformly distributed
random integer.

Command Description Example

randi Generates a single random number between 1 >> a=randi(15)
and imax. a=
9
randi(imax,n) Generates an mxn matrix with random integers >> c=randi(15,3)
between 1 and imax. b=
4 8 11
14 3 8
1 15 8
randi Generates an mxn matrix with random integers >> c=randi(15,2,4)
(imax,m,n) between 1 and imax. c=
1 1 8 13
11 2 2 13
 Randi
• The range of the random integers can be set to be
between any two integers by typing [imin imax]
instead of imax.

>> d=randi([50 90], 3, 4)

d=

83 87 61 89
87 75 72 56
55 53 89 89
 Randn
• The randn command generates normally distributed
numbers with mean 0 and standard deviation of 1.

>> d=randn(3,4)

d=

0.7254 -0.2050 1.4090 -1.2075

-0.0631 -0.1241 1.4172 0.7172
0.7147 1.4897 0.6715 1.6302
 Randn
• The mean and standard deviation of the numbers can be changed
by mathematical operations to have any values.

• Ex) six numbers with a mean of 50 and standard deviation of 6

>> v=6*randn(1,6)+50

v=
51.9556 54.1388 52.9075 48.7862 51.1755 46.8509

>> w=round(6*randn(1,6)+50)

w=
54 45 46 47 38 56
 Exercises
• Define the vectors:
u = −2i + 6j + 5k, v = 5i − 1j + 3k, w = 4i + 7 j − 2k

Use the vectors to verify the identity:

u  ( v  w) = v(u  w) − w(u  v)

Using MATLAB’s built-in functions cross, calculate

the value of the left and right sides of the identity.
 Exercises
• Define the vectors:
u = −2i + 6j + 5k, v = 5i − 1j + 3k, w = 4i + 7 j − 2k

Use the vectors to verify the identity:

u  ( v  w) = v(u  w) − w(u  v)

Using MATLAB’s built-in functions cross, calculate

the value of the left and right sides of the identity.
u=[-2 6 5]; v=[5 -1 3]; w=[4 7 -2];
cross(u,cross(v,w))
dot(u,w)*v-dot(u,v)*w
 Exercises
• Use MATLAB to show that the sum of the infinite

series  1 = 1 + 1 + 1 +  converges to 1.
n
n =1 2 2 22 23

Do it by computing the sum for:

(a) n =10 (b) n=20

(c) n=30 (d) n=40
 Exercises
• Use MATLAB to show that the sum of the infinite series

converges to 1. 1 1 1 1

n =1 2
n
= + 2 + 3 +
2 2 2
Do it by computing the sum for:

(a) n =10 (b) n=20

(c) n=30 (d) n=40

» n=1:10;
» s= sum(1./2.^n)
s=
0.9990