Upload
muhammed-huzaifa
View
230
Download
0
Embed Size (px)
Citation preview
8/12/2019 2E1 LA Notes 1
1/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1 Matrices and determinants
1.1 Matrices
Definition:An mn matrix is a rectangular array of numbers
(m rows and n columns) enclosed in brackets. The numbers
are called theelements of the matrix.
Examples:
(i)A23matrix has 2 rows and 3 columns:
A=
1 2 3
5 6 7
(ii)Heres a33squarematrix:
A= 1 2 35 6 7
8 9 1 0
(iii) Column vectorsare matrices with only one column:
b=
1
5
8
(iv) Row vectorsare matrices which only have one row:
b= (1 2 3) .
Unless specifically stated otherwise, we will assume that vec-
tors are column vectors.
1.1
8/12/2019 2E1 LA Notes 1
2/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
A general real matrix, A Rmn withm nelements is of
the form
A=
a11 a12 a13 . . . a1na21 a22 a23 . . . a2na31 a32 a33 . . . a2n
... ...
... . . . ...
am1 am2 am3 . . . amn
(1)
We refer to the elements via double indices as follows
(i)The first index represents the row.
(ii)The second index represents the column.
Example
a32is the element in row 3, column 2 of the matrix A.
Notation/Conventions:
Use lowercase boldface (or underlined) letters for vectors
a b c (or a, b, c)
Use uppercase boldface (or underlined) letters for matrices
A B C (or A, B, C )
Refer to the respective elements by lowercase letters with the
appropriate number of indices e.g.
bi is a vector element
aij is a matrix element
1.2
8/12/2019 2E1 LA Notes 1
3/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.2 Special matrices
The unit matrix, I, is a square matrix whose only non-zeroelements are on the diagonal and are equal to one, e.g.
I=
1 0 0
0 1 0
0 0 1
, I=
1 0 0 . . . 0 0
0 1 0 . . . 0 0
0 0 1 . . . 0 0...
... ... . . .
... ...
0 0 0 . . . 1 00 0 0 . . . 0 1
All elements of thezero matrix, 0, are equal to zero, e.g.
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 0
... ... ... . . . ... ...0 0 0 . . . 0 0
0 0 0 . . . 0 0
A diagonal matrix only has non-zero elements on the main
diagonal. These non-zero elements can have any value, e.g.
D =
d11 0 00 d22 0
0 0 d33
, D =
d11 0 . . . 0 0
0 d22 . . . 0 0...
... . . . ...
...
0 0 . . . dn1,n1 0
0 0 . . . 0 dnn
are square diagonal matrices.
1.3
8/12/2019 2E1 LA Notes 1
4/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.3 Matrix algebra
1.3.1 Matrix equality
Two matrices areequalif they have the same size and if their
corresponding elements are identical , i.e.
A= B
if and only if
aij =bij fori = 1,...,m; j = 1,...,n
1.3.2 Matrix addition
Two matrices can only be added if they have the same size.
The result is another matrix of the same size.
We add matrices by adding their corresponding elements, i.e.
A= B + C
is obtained (element-wise) via
aij =bij+cij fori = 1,...,m; j = 1,...,n
Example
A= 1 2 3
5 6 7
B=
10 1 235 16 3
A + B=
1 + 10 2 + 1 3 + 23
5 + 5 6 + 16 7 + 3
=
11 3 26
10 22 10
1.4
8/12/2019 2E1 LA Notes 1
5/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.3.3 Multiplication of a matrix by a scalar
A matrix is multiplied by a scalar (= a number) by multiplyingeach element of the matrix by that scalar.
The result is a matrix of the same size.
Hence
A= B
is given (element-wise) by
aij = bij fori = 1,...,m; j = 1,...,n
Example
A=
1 2 3
5 6 7
3A
= 31 32 33
35 36 37 = 3 6 9
15 18 21 1.3.4 Matrix Vector multiplication
Before defining the product of a matrix and a vector, let us
recall the notion of thedot product(or,scalar product) of two
vectors.
Let a Rn and b Rn be two column vectors with n real
elements each:
a=
a1a2a3
...
an
, b=
b1b2b3
...
bn
.
1.5
8/12/2019 2E1 LA Notes 1
6/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
Then thedot product(or,scalar product) ofaandbis defined
as
a b= a1b1+a2b2+a3b3+ +anbn.
Note that the row vector corresponding to the (column) vector
acan be defined by thevector transposeoperation:
a
T
=
a1a2
a3...
an
T
= a1 a2 a3 . . . an .
Then the product of the row vector aT and the column vector
bcan be defined as follows:
aT b def= a b= a1b1+a2b2+a3b3+ +anbn (2)
Example
a =
2
1
3
, b =
5
2
4
; aT b =
2 1 3
5
2
4
=
21
3
52
4
= 10212 =4.
1.6
8/12/2019 2E1 LA Notes 1
7/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
Now we can define a matrix vector multiplication.
The product of an mn matrix Aand a column vector xoflengthn produces a column vector bof lengthm:
Amn
xn1
= bm1
.
This implies that thenumber of columns of the matrixmust
be equal to thenumber of rows in the column vector!
The elements of the vectorb are calculated by taking
dot products of the rows of the matrix with the column vector:
if rows ofA Rmn are vectors aT1 , aT2 ,. . ., a
Tm R
n then
A=
aT1
aT2...
aTm
A x=
aT1 x
aT2 x...
aTm x
=b (3)
whereaTi xare dot products defined in equation (2).
Element wise this isn
j=1
aijxj =bi fori = 1,...,m.
1.7
8/12/2019 2E1 LA Notes 1
8/23
8/12/2019 2E1 LA Notes 1
9/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.3.5 Matrix Vector multiplication using Falks scheme
For practical computations useFalks schemeand evaluate
Ax= b
in tabular form as follows:
x
A b
Example
A=
1 2 3
5 6 7
, x=
32
4
Evaluate Ax as follows:
32
4
1 2 3
5 6 7
Now place the elements of the resulting vector at the inter-
section of the column vector and the rows of the matrix:3
2
4
1 2 3
5 6 7
13 + 22 + 34
53 + 62 + 74 =
19
55
1.9
8/12/2019 2E1 LA Notes 1
10/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.3.6 Matrix Matrix multiplication
The product of an mn matrix A and an np matrix Bproduces anmpmatrix C, i.e.
Amn
Bnp
= Cmp
.
This implies that thenumber of columns of the first matrix
must be equal to thenumber of rows in the second matrix.
Here are some examples:
A=
1 2 3
5 6 7
B=
1 2 35 6 7
8 9 1 0
23 33
C=
1 2 3 25 6 7 2
8 9 1 0 7
D=
1 2
5 6
8 9
3 7
34 42
We can form
AB the result is a23matrix
AC the result is a24matrix
CD the result is a32matrix
It isnotpossible to form
BA or AD
1.10
8/12/2019 2E1 LA Notes 1
11/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
The product matrix of two matrices is obtained by taking dot
products of the rows of the left matrix with the columns of the
right matrix.
If the rows ofA Rmn are the vectors aT1 , aT2 ,. . ., a
Tm R
n
and the columns ofB Rnp are b1, b2,. . ., bp Rn. Let
A=
aT1
a
T
2...
aTm
and B= b1 b2 . . . bp
then the matrix matrix product is
AB=
aT1 b1 aT1 b2 . . . a
T1 bp
aT2 b1 aT2 b2 . . . aT2 bp...
... . . . ...
aTm b1 aTm b2 . . . a
Tm bp
Rmp (4)
Element-wise this is
nj=1
aijbjk =cik fori = 1,...,m; k= 1,...,p.
1.11
8/12/2019 2E1 LA Notes 1
12/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
Example
A= 1 23 4
B= 4 3 2
2 1 1
The respective row and column matrices are
aT1 = (1 2)
aT2 = (3 4)b1=
4
2
, b2=
3
1
, b3=
2
1
,
the matrix-matrix product is then
AB=
(1 2)
4
2
(1 2)
3
1
(1 2)
2
1
(3 4)
4
2
(3 4)
3
1
(3 4)
2
1
which can be evaluated by a series of dot products. That is
AB=
14 + 22 13 + 21 12 + 21
34 + 42 33 + 41 32 + 41
Giving the final result
C= 8 5 4
20 13 10
1.12
8/12/2019 2E1 LA Notes 1
13/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.3.7 Matrix Matrix multiplication using Falks scheme
For practical computation we again use Falks scheme and
evaluate AB= C in tabular form as follows:
B
A C
Example
A= 1 2
3 4
, B=
4 3 22 1 1
Evaluate AB as follows
4 3 2
2 1 1
1 2 (14 + 22) (13 + 21) (12 + 21)
3 4 (34 + 42) (33 + 41) (32 + 41)
Place the elements at the intersection of the rows of the left
matrix and the columns of the right matrix
The result
4 3 2
2 1 1
1 2 ( 14 + 22) (13 + 21) (12 + 21)
3 4 ( 34 + 42) (33 + 41) (32 + 41)
therefore gives
AB=
8 5 4
20 13 10
1.13
8/12/2019 2E1 LA Notes 1
14/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.3.8 Differences from multiplication with numbers
(i)Matrix multiplication isnot commutative
AB= BA
Example
1 2
4 1
2 2
3 1
=
12 + 23 12 + 21
42 + 13 42 + 11
=
8 4
11 9
2 2
3 1
1 2
4 1
=
21 + 24 22 + 21
31 + 14 32 + 11
= 10 67 7
We must be careful how we multiply out!
(ii) AB= 0 does not imply A= 0, B= 0 or BA= 0
Example 1 1
2 2
1 1
1 1
=
0 0
0 0
1 1
1 1
1 1
2 2
=
1 1
1 1
1.14
8/12/2019 2E1 LA Notes 1
15/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
(iii) AC= ADdoes not necessarily imply C= D
Example 1 1
2 2
2 1
2 2
=
4 3
8 6
1 1
2 2
3 0
1 3
=
4 3
8 6
(iv) BUT other properties are similar to numbers
A(B + C) = AB + AC distributive law
A(BC) = (AB)C associative law
1.15
8/12/2019 2E1 LA Notes 1
16/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.4 Transpose of a matrix
The transpose of a matrix is obtained by interchanging itsrows and columns
aTij =aji fori = 1,...,m; j = 1,...,n
The transpose is denoted by a superscript T and the general
matrix given in equation (1) becomes
AT =
a11 a21 a31 . . . am1a12 a22 a32 . . . am2a13 a23 a33 . . . am3
... ...
... . . . ...
a1n a2n a3n . . . amn
Example
A= 1 2 3
5 6 7
A
T = 1 52 6
3 7
IfA= AT then A is asymmetric matrix, e.g.
A=
3 2 1
2 7 01 0 8
The matrix transpose also satisfies the followingrules:
i)(AT)T = A for any matrix A;
ii)(A + B)T = AT + BT and (AB)T = BTAT,
provided that matrices A and B have compatible dimensions.
1.16
8/12/2019 2E1 LA Notes 1
17/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
< Proof of the identity (AB)T = BTAT
If the rows ofA Rmn are the vectors aT1 , aT2 ,. . ., aTm Rnand the columns ofB Rnp are b1, b2,. . ., bp R
n.
Taking the transpose of equation (4)
(AB)T =
aT1 b1 a
T2 b1 . . . a
Tm b1
aT1 b2 aT2 b2 . . . a
Tm bp
.
.. .
.. .
. . .
..aT1 bp a
T2 bp . . . a
Tm bp
For column vectors ai and bj vector multiplication a
Ti bj is
defined as the dot product betweenaiand bj. The dot product
is commutative so
aTi bj =ai bj =bj aj =bT
j ai
which implies that
(AB)T =
bT1 a1 bT1 a2 . . . b
T1 am
bT2 a1 bT2 a2 . . . b
T2 am
... ... . . .
...
bTp a1 bT
p a2 . . . bT
p am
= BTAT
since
BT =
bT1bT2
...
bTp
and AT =
a1 a2 . . . am
. >
Here and further in these notes, the material between the
markers < and > is for advanced reading.1.17
8/12/2019 2E1 LA Notes 1
18/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.5 Determinant of a matrix
The determinant of a22matrix
A=
a11 a12a21 a22
is writtendet A or|A|or
a11 a12a21 a22
=
a11a22a12a21
Example
A=
1 2
4 7
, det A= 78 =15.
The determinant of a33matrix is written as
|A|=
a11 a12 a13a21 a22 a23a31 a32 a33
=a11
a22 a23a32 a33a12
a21 a23a31 a33 +a13
a21 a22a31 a32
=a11(a22a33a32a23)a12(a21a33a31a23)+ a13(a21a32a31a22)
That is the33 determinant is defined in terms of determi-
nants of2 2sub-matrices ofA. These are called theminors
ofA.
1.18
8/12/2019 2E1 LA Notes 1
19/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
Example
m12=
a11 a12 a13a21 a22 a23a31 a32 a33
= a21 a23a31 a33
is obtained by suppressing the elements in row 1 and column
2 of matrix A.
Cofactors
Thecofactorcij is defined as the coefficient ofaij in the de-
terminantA. If is given by the formula
cij = (1)i+jmij
where theminoris the determinant of order(n 1) (n 1)
formed by deleting the column and row containing aij.Examples
c11= (1)1+1m11= +1
a11 a12 a13a21 a22 a23a31 a32 a33
=a22a33a32a23
c23= (1)2+3m23=1
a11 a12 a13a21 a22 a23a31 a32 a33
=a11a32+ a31a12
1.19
8/12/2019 2E1 LA Notes 1
20/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
General determinant
The value of annndeterminant equals the sum of the prod-ucts of the elements in any row (or column) and their cofac-
tors, i.e.
|A|=n
j=1
aijcij, for i= 1, . . . , n1, orn
or
|A|=
ni=1
aijcij, for j = 1, . . . , n1, orn
Example
For a33matrix
det A=a11c11+a12c12+a13c13 (1st row)
or
det A= a12c12+a22c22+a32c32 (2nd column)
Points to note:
the determinantdet A is equal to zero if
(i)rows or columns ofA are multiples of each other,
(ii)rows or columns are linear combinations of each other,
(iii)entire rows or columns are zero;
ifdet A= 0the matrix A is called a singular matrix;
for any square matrices A and B there holds
det A= det(AT), det(AB) = det(A)det(B).
for the unit matrix I one hasdet I= 1.
1.20
8/12/2019 2E1 LA Notes 1
21/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
1.6 The matrix inverse
The inversea1
of a scalar (=a number)a is defined by
a a1 = 1.
For square matrices we use a similar definition: the inverse
A1 of annmatrix A fulfils the relation
AA1 = I
where Iis thenn unit matrix defined earlier.
Note: ifA1 exists then
det(A)det(A1) = det(AA1) = det I= 1.
Hence,det(A1) = (det A)1.
Example
The inverse of
A= 3 27 5 is given by
B= A1 =
5 2
7 3
since
AB=
5 2
7 33 2 ( 3527) (32 + 23)
7 5 ( 7557) (72 + 53)
which gives
AB=
1 0
0 1
= I
as required.
1.21
8/12/2019 2E1 LA Notes 1
22/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
The matrix inverse can be computed as follows
1.Find the determinantdet A
2.Find the cofactors of all elements in A and form a new
matrix Cof cofactors, where each element is replaced by
its cofactor.
3.The inverse ofA is now given as
A1
= C
T
det A
Note: the inverse A1 exists if (and only if)det A= 0.
ExampleFind the inverse of
A=
1 1 23 1 23 2 1
.
det A= 1
1 2
2 1
(1)
3 2
3 1
+ 2
3 1
3 2
= 13 + 1(3) + 23
= 6
Since the determinant is nonzero an inverse exists.
1.22
8/12/2019 2E1 LA Notes 1
23/23
2E1: Linear Algebra|Lecture Notes 1 Matrices and matrix algebra
Calculate the matrix of minors
M=
1 22 1 3 23 1 3 13 21 22 11 23 1
1 13 2
1 2
1 2
1 2
3 2
1 1
3 1
= 3 3 35 7 14 8 2
Modify the signs according to whetheri+j is even or odd to
calculate the matrix of cofactors
C=
3 3 3
5 7 14 8 2 .
It follows that
A1 =
1
6C
T =1
6
3 5 43 7 8
3 1 2
.
To check that we have made no mistake we can compute
A1
A=1
6
3 5 43 7 8
3 1 2
1 1 23 1 2
3 2 1
=
1 0 00 1 0
0 0 1
.
This way of computing the inverse is only useful for hand
calculations in the cases of22or33matrices.
1.23