(The Transpose Operator)

>> C=[2 55 14 8; 21 5 32 11; 41 64 9 1]

C = 2 55 14 8

21 5 32 11

41 64 9 1

>> D=C'

D = 2 21 41

55 5 64

14 32 9

8 11 1
1
 (Vectors Arithmetic 2)
 Addition or subtraction of matrices is possible as long as they
are the same size.
>> [1 2 3]+[4 5 6]
ans = 5 7 9

 To deal with arrays on an element-by-element level we need to use

the following array or dot-operators:
.* , ./ and .^

 Element-by-element multiplication and division uses the .*

and ./ operators. The * and / operators are used for complete
matrix multiplication and division.
>> [1 2 3].*[4 5 6]
ans = 4 10 18
2
 dot-operators
.* , ./ and .^
Because scalars are equivalent to a 1×1 array, you can
either use the standard or the dot-operators when doing
multiplication, division and exponentiation of scalars (i.e.,
of single numbers).
It is okay for you to always use the dot-operators, unless
you intend to perform vector or matrix multiplication or
division.
Array vs. Matrix Operations
 Example: m=[2,1;3,4] p=[5,6;7,8] (in class we had a look on
m=[1,2;3,4] and results were different )

q=m.*p

results in [10, 6; 21, 32]; this is array multiplication

q = m * p
results in [17, 20; 43, 50]; this is matrix multiplication

So, do NOT forget the dot when doing array operations!

(.* ./ .^)
 (More about Column Vectors)
 The vectors seen so far are row vectors (horizontal).
You can define column vectors as well. For row vectors
you separate the elements with spaces or commas; for
column vectors you use semi-colons.
>> colvec = [1;2;3]
colvec = 1
2
3

5
 (More about Matrices)
 You can define matrices the same way by combining rows
and columns.
>> mat1 = [1 2;3 4]
mat 1 = 1 2
3 4
>> mat2 = [mat1;mat1]
mat2 = 1 2
3 4
1 2
3 4

6
 (The .^ and sqrt Commands)
 .^ us used for the element-by-element power operator
(works like .* and ./ seen in slide 25).
>> [1 2 3].^2
ans = 1 4 9
 Square roots are done with sqrt (works with negative
numbers too!).
>> sqrt (-9)
ans = 0 + 3.0000i
>> sqrt ([1 2 4 9])
ans = 1.0000 1.4142 2.0000 3.0000

7
 (Trig Commands)
 Trig functions are supported like sin.
>> sin ([pi/4 pi/2 pi])
ans = 0.7071 1.0000 0.0000
 A few more command examples... There are many many
more!
 log(x), log10(x), cos(x), atan(x), exp(x), round(x),
floor(x), ceil(x), angle(x), abs(x)...
 They work exactly like the math library functions in C.
Need help with the sin function? Try help sin or doc sin.

8
(Saving and Loading Variables)
 Use save to save variables to a file.
>> save myfile mat1 mat2
saves matrices mat1 and mat2 to the file myfile.mat

 To load saved variables back into the workspace, use

load.
>> load myfile

 If the command window is too full, use clc.

9
 (More about Transpose)
 The transpose operators turn a column vector into a row
vector and vice versa as seen before. For example to
transpose a vector x,

>> transpose (x);

or
>> x'

 You can also use x.' if you want to transpose a vector

containing complex numbers.

10
 (Example: Scalar Product)
 A scalar product can be done with one row vector
by a column vector. A transpose may be required.

>> scalarprod = [1 2 3 4] * [4 5 6 7]'

scalarprod = 60

11
(Multiple Initializations)
 It is possible to perform multiple initializations at once:

>> v1 = ones (1,10)

v1 = 1 1 1 1 1 1 1 1 1 1

>> v1 = rand (1,100);

A row vector with 100 random elements between 0 and 1.

>> v1 = zeros (45,1);

A column vector with 45 zero elements.

12
 (Index vs. Subscript)
 For a vector, indexes and subscripts are the same.
Remember that indexes and subscripts start at 1 in
MATLAB.
 For matrices, subscripts have the row and column
numbers separated by a comma. Indexes, on the other
hand, indicate the linear position from the start of the
matrix.
subscripts indexes

13
 (Matrix Indexing)
 The index argument can be a matrix. In this case,
each element is looked up individually, and
returned as a matrix of the same size as the index
matrix.

>> a = [10 20 30 40];

>> b = a([1 2 4;3 4 2])
b = 10 20 40
30 40 20

14
(Working Whole Rows or Columns)
 The colon operator selects rows or columns of a matrix.
>> c = b (1,:)
c = 10 20 40

>> d = b (:,2)
d = 20 40

>> b (:, 3) = [100 200]

b = 10 20 100
30 40 200

15
 (Colon Operator)
The colon : lets you address a range of elements

• Vector (row or column)

 va(:) - all elements

 va(m:n) - elements m through n

• Matrix

 A(:,n) - all rows of column n

 A(m,:) - all columns of row m

 A(:,m:n) - all rows of columns m through n

 A(m:n,:) - all columns of rows m through n

 A(m:n,p:q) - columns p through q of rows m through n

16
17
 (Colon Operator)
The colon : lets you address a range of elements

• Vector (row or column)

 va(:) - all elements

 va(m:n) - elements m through n

• Matrix

 A(:,n) - all rows of column n

 A(m,:) - all columns of row m

 A(:,m:n) - all rows of columns m through n

 A(m:n,:) - all columns of rows m through n

 A(m:n,p:q) - columns p through q of rows m through n

18
 Colon Operator
 Colon notation can be used to define evenly spaced vectors in the form:
first : last mm=1:3 mm=
1 2 3

 The default spacing is 1. To use a different increment use the form:

first : increment : last y=1:2:7 y=

1 3 5 7

 The numbers now increment by 2

19
 Extracting Data
 The colon represents an entire row or column when used in as an array
index in place of a particular number.

n = n(:,1) n(2,:)
2 50 20 ans =
1 1 2 ans = 1 1 2
2
0 55 66
1
The colon operator can also be used to extract 0
a range of rows or columns:

n(2:3,:) n(1,2:3)
ans = ans =
1 1 2 50 20
0 55 66

20
(Finding Minimum and Maximum Values)
>> vec = [10 7 8 13 11 29];
 To get the minimum value and its index:

>> [minVal,minInd] = min(vec)

minVal = 7 minInd = 2
 To get the maximum value and its index:

>> [maxVal,maxInd] = max(vec)

maxVal = 29 maxInd = 6
>> [a,b] = max(vec)
a = 29 b = 6

21
 (Getting Specific Columns)


You can replace vector index or matrix
indices by vectors in order to pick out
specific elements. For example, for
vector v and matrix m
• v([a b c:d]) returns elements a, b, and
c through d
• m([a b],[c:d e]) returns columns c
through d and column e of rows a and b

22
 (Colon Operator Examples 1)

>> v=4:3:34
v = 4 7 10 13 16 19 22 25 28 31 34

>> u=v([3, 5, 7:10])

u = 10 16 22 25 28 31

>> u=v([3 5 7:10])

u = 10 16 22 25 28 31
23
 (Colon Operator Examples 2)

>> A=[10:-1:4; ones(1,7); 2:2:14; zeros(1,7)]

A = 10 9 8 7 6 5 4

1 1 1 1 1 1 1

2 4 6 8 10 12 14

0 0 0 0 0 0 0

>> B=A([1 3],[1 3 5:7])

B = 10 8 6 5 4

2 6 10 12 14

24
 Example?
A=
10 9 8 7 6 5 4
1 1 1 1 1 1 1
2 4 6 8 10 12 14
0 0 0 0 0 0 0

B=A([1 3],:)
B= B=A([1 3])
10 9 8 7 6 5 4 B=
2 4 6 8 10 12 14 10 2

B=A(:,[1 3 5:7])
B=A([1 3],[1 3 5:7])
B=
B=
10 8 6 5 4
10 8 6 5 4
1 1 1 1 1
2 6 10 12 14
2 6 10 12 14
0 0 0 0 0

25
(Two ways to add elements to existing variables)
Assign values to indices that don't exist:
MATLAB expands array to include indices, puts the specified
values in the assigned elements, fills any unassigned new
elements with zeros.
Add values to ends of variables:
Adding to ends of variables is called appending or
concatenating.
"end" of vector is right side of row vector or bottom of
column vector.
"end" of matrix is right column and bottom row.
26
 (Assigning to undefined indices of vectors)
>> fred=1:4

fred = 1 2 3 4

>> fred(5:10)=10:5:35

fred = 1 2 3 4 10 15 20 25 30 35

>> wilma=[5 7 2]

wilma = 5 7 2 New elements added

>> wilma(8)=4

wilma = 5 7 2 0 0 0 0 4

>> barney(5)=24

barney = 0 0 0 0 24 Unassigned elements

27 are initialized at zero
 (Appending to vectors)
• You can only append row vectors to row vectors and
column vectors to column vectors.
If r1 and r2 are any row vectors,
r3 = [r1 r2] is a row vector whose left part
is r1 and right part is r2
If c1 and c2 are any column vectors,
c3 = [c1; c2] is a column vector whose top
part is c1 and bottom part is c2

28
 >> pebbles = [3 8 1 24];

>> dino = 4:3:16;

>> betty = [pebbles dino]

betty = 3 8 1 24 4 7 10 13 16

>> bedrock = [pebbles'; dino']

bedrock = 3
8
1
24
4
7
10
13
29
16
 (Appending to matrices)
• If appending one matrix to the right side of
another matrix, both must have same
number of rows.
• If appending one matrix to the bottom of
another matrix, both must have same
number of columns

30
 >> A2=[1 2 3; 4 5 6]
A2 = 1 2 3
4 5 6
>> B2=[7 8; 9 10]
B2 = 7 8
9 10
>> C2=eye(3)
C2 = 1 0 0
0 1 0
0 0 1

31
 >> Z=[A2 B2]

Z = 1 2 3 7 8

4 5 6 9 10

>> Z=[A2; C2]

Z = 1 2 3

4 5 6

1 0 0

0 1 0

0 0 1

>> Z=[A2; B2]

??? Error using ==> vertcat

CAT arguments dimensions are not consistent.

32
To delete elements in a vector or matrix, set range
to be deleted to empty brackets.
>> kt=[2 8 40 65 3 55 23 15 75 80]
kt = 2 8 40 65 3 55 23 15 75 80
>> kt(6)=[] Delete sixth element (55)

kt = 2 8 40 65 3 23 15 75 80 55 gone

>> kt(3:6)=[] Delete elements 3 through

6 of current kt, not
original kt
kt = 2 8 15 75 80

33
 To delete elements in a vector or matrix, set
range to be deleted to empty brackets.
>> mtr=[5 78 4 24 9; 4 0 36 60 12; 56 13 5 89
3]

mtr = 5 78 4 24 9

4 0 36 60 12

56 13 5 89 3

>> mtr(:,2:4)=[] Delete all rows for

columns 2 to 4

mtr = 5 9

4 12

34 56 3
MATLAB has many built-in functions for working
with arrays. Some common ones are:
• length(v) - number of elements in a vector.
• size(A) - number of rows and columns in a
matrix or vector.
• reshape(A,m,n) - changes number of rows
and columns of a matrix or vector while keeping
total number of elements the same. For example,
changes 4x4 matrix to 2x8 matrix.

35
 reshape Example
>> a = [1 2 ; 3 4]
a=
12
34

>> a = [a [5 6; 7 8]]
a= ans =
1256 1
3478
3
>> reshape (a, 1, 8) 2
ans = 4
13245768 5
7
>> reshape (a, 8, 1)
6
8
36
 • diag(v) - makes a square matrix of zeroes with
vector in main diagonal.
• diag(A) - creates vector equal to main diagonal
of matrix.

For more functions, in help window select

"Functions by Category", then
"Mathematics", then "Arrays and Matrices".

37
 Example$

>> a=[1 2 3]
>> B= [1 2 3;2 5 7; 0 0 8]
a=
B=
1 2 3
1 2 3
2 5 7
>> diag(a)
0 0 8
ans =
>> diag(B)
1 0 0
ans =
0 2 0
0 0 3
1
5
8
38