LINEAR ALGEBRA & VECTOR ANALYSIS

 Lecture 4 #: Section 2.6_2.7

Mr. Ibrahima Gueye 2/28/2023 1

Mathematics Faculty
OBJECTIVES

 Factorize matrix 𝐴 = 𝐿𝑈, 𝑎𝑛𝑑 𝐴 = 𝐿𝐷𝑈

 Compute the transpose of a Matrix 𝐴𝑇
 Use permutation matrix to exchange rows
 Symmetric Matrix

2/28/2023 2
MATRIX FACTORIZATION

 Matrix factorization is the process of writing A into a product of two matrices

(𝐿, 𝑡ℎ𝑒 𝑙𝑜𝑤𝑒𝑟 𝑡𝑟𝑖𝑎𝑛𝑔𝑢𝑙𝑎𝑟 𝑚𝑎𝑡𝑟𝑖𝑥, 𝑎𝑛𝑑 𝑈, 𝑡ℎ𝑒 𝑢𝑝𝑝𝑒𝑟 𝑡𝑟𝑖𝑎𝑛𝑔𝑢𝑙𝑎𝑟 𝑚𝑎𝑡𝑟𝑖𝑥)
 𝐴 = 𝐿𝑈
 From previous lecture, we used elimination matrices to go from A to U (upper triangular)
 𝐸32 𝐸31 𝐸21 A = U (1)
−1 −1 −1
 We shall multiply both side of Equation (1) by (𝐸21 𝐸31 𝐸32 ) to recover A.
−1 −1 −1 −1 −1 −1 −1 −1 −1
 𝐸21 𝐸31 𝐸32 𝐸32 𝐸31 𝐸21 𝐴 = 𝐸21 𝐸31 𝐸32 𝑈 where 𝐿 = 𝐸21 𝐸31 𝐸32
Hence, 𝐴 = 𝐿𝑈

2/28/2023 3
EXAMPLE 1(2X2)

 Factorize the (2x2) matrix A into LU

2 1
 𝐴=
8 7

2/28/2023 4
SOLUTION

2 1
 𝐴=
8 7
 We will use Gaussian Elimination using elementary matrix
1 0
 𝐸21 =
−4 1
1 0 2 1 2 1
 𝐸21 𝐴 = = =𝑈
−4 1 8 7 0 3
−1 1 0 2 1
 𝐸21 𝐸21 𝐴 = ∙ = 𝐿𝑈
+4 1 0 3
 Hence, 𝐴 = 𝐿𝑈

2/28/2023 5
EXAMPLE 2

 Factorize the (3x3) matrix into A=LU

2 1 0
 𝐴 = −1 2 1
1 0 3

2/28/2023 6
SOLUTION
2 1 0 1 0 0
1
 𝐴 = −1 2 1 𝐸21 = 2 1 0 We shall multiply 𝐸21 𝑏𝑦 𝐴
1 0 3 0 0 1
1 0 0 2 1 0 2 1 0 1 0 0
1 5
 𝐸21 𝐴 =
2
1 0 ∙ −1 2 1 = 0 2
1 𝐸31 = 01 1 0 𝑊𝑒 𝑠ℎ𝑎𝑙𝑙 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑦 𝐸31 𝑏𝑦 𝐸21 𝐴
0 0 1 1 0 3 1 0 3 − 0 1
2

1 0 0 2 1 0 2 1 0 1 0 0
5
 𝐸31 𝐸21 𝐴 = 0 1 0 ∙ 0
5
1 = 0 2
1 𝐸32 = 0 11 0
1 2 1
−2 0 1 1 0 3 0 −2 3 0 1
5

1 0 0 2 1 0 2 1 0
5 5
 𝐸32 𝐸31 𝐸21 𝐴 = 0 1 0 ∙ 0 1 = 0 2
1 = 𝑈 𝑎𝑛𝑑 𝐿 = 𝐸 −1 𝐸 −1 𝐸 −1
2 21 31 32
1 1 16
0 5 1 0 −2 3 0 0 5

2/28/2023 7
SOLUTION CONTINUE

−1 −1 −1
 𝐿 = 𝐸21 𝐸31 𝐸32

1 0 0 1 0 0 1 0 0
−1 −1
 𝐸31 𝐸32 = 01 1 0 ∙ 0 1
−1
0 = 0 1
−1
0
+ 0 1 0 1 1/2 1
2 5 5

1 0 0 1 0 0 1 0 0
1
−1
 𝐸21 −1 −1
(𝐸31 𝐸32 ) =
−1
1 0 ∙ 01 1 0 = −2 1 0 =𝐿 𝑓𝑜𝑟 𝑙𝑜𝑤𝑒𝑟 𝑡𝑟𝑖𝑎𝑛𝑔𝑢𝑙𝑎𝑟 𝑚𝑎𝑡𝑟𝑖𝑥
2 −1 1 −1
0 0 1 2 5
1 1
2 5
 Hence, 𝐴 = 𝐿𝑈
1 0 0 2 1 0
2 1 0 1 5
 −1 2 1 = − 2
1 0 ∙ 0 1
2
1 −1 16
1 0 3 1 0 0
2 5 5

2/28/2023 8
FACTORIZE A=LDU

2 1 0
 𝐴 = −1 2 1
1 0 3
 𝐴 = 𝐿𝑈

1 0 0 2 1 0
2 1 0 1 5
− 1 0 ∙ 0 1
 −1 2 1 = 2 2
1 −1 16
1 0 3 1 0 0
2 5 5
 D is the diagonal pivot matrix generated from the diagonal entries of U (the upper triangular matrix)

1 0 0 2 0 0 1
1
0
1 5 2
 𝐴 = 𝐿𝐷𝑈 = − 1 0 ∙ 0 0 2
2 2 0 1
1 −1 16 5
1 0 0
2 5 5 0 0 1
2/28/2023 9
EXAMPLE 3

 Factorize the following matrix A into LDU

1 0 1
 𝐴= 2 2 2
3 4 5

2/28/2023 10
MATRIX TRANSPOSE

 Given an (mxn) matrix A, its transpose is 𝐴𝑇 and its size is (nxm). The rows of A are the Columns of 𝐴𝑇 and vice
versa.
2 −3 4
 Example: 𝐴 = 𝐴 𝑖𝑠 𝑎 2 𝑏𝑦 3 𝑚𝑎𝑡𝑟𝑖𝑥
1 0 5
2 1
 𝐴𝑇 = −3 0 (𝐴𝑇 𝑖𝑠 𝑎 3 𝑏𝑦 2 𝑚𝑎𝑡𝑟𝑖𝑥
4 5

2/28/2023 11
OPERATIONS WITH TRANSPOSE

 Sum of 𝐴 + 𝐵 𝑖𝑠 𝐴𝑇 + 𝐵𝑇
 Product of AB is (𝐴𝐵)𝑇 = 𝐵𝑇 𝐴𝑇
 Inverse the transpose of 𝐴−1 𝑖𝑠 (𝐴−1 )𝑇 = (𝐴𝑇 )−1
 If 𝐴 = 𝐿𝐷𝑈, 𝑡ℎ𝑒𝑛 𝐴𝑇 = (𝐿𝐷𝑈)𝑇 = 𝑈 𝑇 𝐷 𝑇 𝐿𝑇

2/28/2023 12
SYMMETRIC MATRIX
 Definition:
A symmetric matrix has 𝐴𝑇 = 𝐴. This means that 𝑎𝑗𝑖 = 𝑎𝑖𝑗
 Example:
1 2
 𝐴=
2 5
1 2
 𝐴𝑇 =
2 5
 If A is a symmetric matrix, 𝐴 = 𝐿𝐷𝑈 𝑤𝑖𝑙𝑙 𝑙𝑒𝑎𝑑 𝑡𝑜 𝐴 = 𝐿𝐷𝐿𝑇

2/28/2023 13
EXAMPLE

−1 0
−1 1 0
 Consider the rectangular matrix 𝑅 = , 𝑎𝑛𝑑 𝑅𝑇 = 1 −1
0 −1 1
0 1
−1 0
−1 1 0 2 −1
 𝑅𝑅 𝑇 = ∙ 1 −1 = 𝑖𝑠 𝑎 𝑠𝑦𝑚𝑚𝑒𝑡𝑟𝑖𝑐 𝑚𝑎𝑡𝑟𝑖𝑥
0 −1 1 −1 2
0 1
−1 0 1 −1 0
−1 1 0
 𝑅𝑇 𝑅 = 1 −1 ∙ = −1 2 −1 𝑖𝑠 𝑎𝑙𝑠𝑜 𝑎 𝑠𝑦𝑚𝑚𝑒𝑡𝑟𝑖𝑐 𝑚𝑎𝑡𝑟𝑖𝑥
0 −1 1
0 1 0 −1 1

2/28/2023 14
PERMUTATION MATRICES
Definition:
A permutation matrix P has the rows of the identity matrix I in any order
1 0 0 0 1 0 0 0 1 1 0 0
𝐼= 0 1 0 𝑃21 = 1 0 0 𝑃31 = 0 1 0 𝑃32 = 0 0 1
0 0 1 0 0 1 1 0 0 0 1 0

1 0 0 0 1 0 0 1 0
𝑃32 𝑃21 = 0 0 1 1 0 0 = 0 0 1
0 1 0 0 0 1 1 0 0

0 1 0 1 0 0 0 0 1
𝑃21 𝑃32 = 1 0 0 0 0 1 = 1 0 0
0 0 1 0 1 0 0 1 0
𝑃−1 = 𝑃𝑇

2/28/2023 15
MATRIX FACTORIZATION WHERE ROW EXCHANGE IS NEEDED

 If A is invertible, a permutation P will put its rows in the right order to factor 𝑃𝐴 = 𝐿𝑈.
 There must be a full set of pivots after row exchanges for A to be invertible.

2/28/2023 16
EXERCISES

2/28/2023 17
EXERCISES

2/28/2023 18
CONCLUSION

 Practice Makes Perfect!!!

2/28/2023 19