Diagonalization

of a matrix

Mathematics report CA1

Sneha sarkar, B-127

Mrs.Soumendu Golui
Professor of Mathematics
Department of Mathematics
Silliguri Institute of Technology
01
INDEX
S.L. No. Name of the topic Page No.
1. Introduction 3
2. Some important terms 4-7
3. Matrix diagonalization 8
4. How to Diagonalize a Matrix 8
5. Diagonalizable Matrix 8-9
6. Non Diagonalizable Matrix 9
7. Application 10
8. Conclusion 10
9. References 10

02
Introduction

Diagonalizable matrices are of the interest because these are especially easy to handle. One can raise a
diagonal matrix to a power by simply rising the diagonal entries to that same power and the determinant of a
diagonal matrix is simply the product of all diagonal entries. Geometrically diagonalizable matrix is an
inhomogeneous dilation- it scales the space as does a homogeneous dilation.
The primary motive of this report is to understand the concept of diagonalization, to understand its
applications in real world scenarios. Matrices play various role in linear algebra, physics , computer science,
business studies, engineering etc.
This report delves into the intricacies of diagonalization, examining, real-world applications, and concluding
with a comprehensive conclusion, outlining key takeaways and potential future research directions.

03
Some Important Terms
1)Algebraic Multiplicity and Geometric Multiplicity of Eigenvalues:
If an eigenvalue λ repeats and repeats r times in a characteristic equation then we say the eigen value λ has algebraic multiplicity r. On the
other hand, the number of independent eigen vectors corresponding to an eigen value λ is called its geometric multiplicity. That is
geometric multiplicity of = λ dimension of eigen space of with it, essentially denoting the dimension of the eigen space of λ.

3 1 −1
Example- Find the Algebraic Multiplicity and Geometric Multiplicity of the matrix A= 2 2 −1
2 2 0
3−λ 1 −1
Characterististic equation of the given matrix A is 2 2 − λ −1 =0
2 2 0−λ
=>𝜆3 − 5𝜆2 + 8𝜆 − 4 = 0
3
=>(𝜆 − 1) 𝜆 − 2 =0
This is show that 𝜆= 1,2,2
Because the eigen value 2 repeats 2 times . therefore, the algebraic multiplicity of the eigen value 2 is 2 .
Corresponding to 𝜆= 2 if α=(x1,x2,x3) be eigen vector then
Aα= 2α => Aα= 2Iα => (A-2I)α = 0

04
 Therefore, the number of independent eigen vectors
That is, x1 + x2 - x3 =0 x1 + x2 - x3 =0
corresponding to the eigen value 𝜆= 2 is 1
2x1 - x3 = 0 => 2x1 - x3 = 0
The geometric multiplicity of the eigen value 2 is 1
2x1 + 2x2 - 2x3 =0 x1 + x2 - x3 =0 ………………………..(1)
The algebraic multiplicity of the eigen value 1 is 1

solution of (1) give the eigen vector

1 1 −1
its coefficient matrix is 2 0 −1 2) Regular Eigen Values:
2 2 −2
An eigenvalue is deemed regular when its
applying elementary row operation on it algebraic multiplicity matches its geometric
multiplicity indicating a balance between
1 1 −1 1 1 −1
𝑅2−2𝑅1 the two aspects of its existence in the matrix.
2 0 −1 𝑅3−𝑅1 > 0 −2 1
2 2 −2 0 0 0
which is the Echelon matrix having two non-zero rows , Example- In the previous example the eigen
therefore rank of the coefficient is 2. value 2 is not regular.

Now number of independent solution of (1)

= Number of unknown – Rank of the coefficient matrix =3-2=1

05
3)Eigen Values:
4)Eigen vector:
To find eigen values first we find the characteristic equation given by An eigen vector of a square matrix A is a non zero vector X such
A- 𝜆I = 0 that for some number 𝜆, We have,
AX = 𝜆X
Now the roots of the equation or values of 𝜆 are called eigen values AX - 𝜆𝐼X =0
Example- ( A - 𝜆𝐼) x =0
where 0 represents the zero vector,
Let A be a square matrix of order 3x3
1 1 3 Example
1 5 1
3 1 1 Now eigen vector corresponding to
According to Cayley Hamilton theorem A- 𝜆I = 0 𝜆= -2,3,6
For 𝜆=-2
where 𝜆 is any scalar and, I is the 3x3 unit matrix 3x1 + x2 +3x3 =0
x1 +7 x2 +x3 =0
3x1 + x2 +3x3 =0
The determinant matrix equated to zero
1−𝜆 1 3 By solving these equations , we get
1 5−𝜆 1 =0 x1 =1 , x2=0, x3=-1
3 1 1−𝜆
For 𝜆=3
on expanding the determinant , we get the equation 2x1 + x2 +3x3 =0
𝜆3 − 7𝜆2 + 36 = 0 x1 +2x2 +x3 =0
on solving these equations we get, 𝜆=-2,3,6 these are called 3x1 + x2 -2x3 =0
By solving these equation we get
eigen values. x1 =1 , x2=-1, x3=1
06
For 𝜆=6
-5x1 + x2 +3x3 =0 6) Scalar matrix:
x1 +x2 +x3 =0
3x1 + x2-5x3 =0
If all the diagonal elements of a matrix are same then
By solving these equations , we get the diagonal matrix is scalar matrix
x1 =1 , x2=2, x3=1
4 0 0
Hence Example- 0 4 0
𝜆= -2(1,0,-1) 0 0 4
𝜆= 3(1,-1,1)
𝜆= 6(1,2,1)

Hence corresponding to each eigen value there

exist an eigen vector. 7) Similar Matrices:

5) Diagonal Matrix: Matrices A and B are considered similar if there exists a

non-singular square matrix P of the same order
such that B = 𝑃−1 AP. This relation highlights the
A square matrix whose all elements other than the diagonal equivalence of the two matrices under a change of
elements are 0 is called diagonal matrix. basis, emphasizing the importance of the transformation
matrix P.
4 0 0
6 0
Example- 0 −7 0 and are diagonal square matrices
0 9
0 0 3
07
Matrix Diagonalization solutions. However, a 3x3 matrix has a characteristic polynomial
equation of the third degree that has three solutions. To
Two matrices, A and D, are said to be similar if, for any know how to diagonalize a matrix follow the three steps:
matrix P, the equation: D = P¯¹AP or the equation: A = PDP¯¹ Step 1: Find the matrix determinant and trace.
remains true. Similar matrices share the same eigenvalues. Step 2: Write the characteristic polynomial equation and set it
They also have the same trace and the same determinant. to zero, then find its solutions.
Step 3: Write the diagonal matrix
Hence, matrix D is called the diagonal representation of matrix
A only if there is an invertible matrix P such that D = P¯¹AP
and matrix A is then called a diagonalizable matrix. Diagonalizable Matrix
Let A be a 2x2 square matrix in the general form
𝑎 𝑏 A matrix is diagonalizable when it has distinct eigenvalues. This
A= that has two eigenvalues λ₁, λ₂, then
𝑐 𝑑 fact reflects in the characteristic polynomial equation by
the diagonal matrix D is of the form:
having solutions.
𝝀₁ 0
D= , therefore, diagonalization requires finding For Example:
0 𝝀₂
1 2
the eigenvalues of the matrix. For the matrix find its eigenvalues and prove that it
3 2
is diagonalizable.
How to Diagonalize a Matrix Solution:
Step 1: find the matrix trace and determinant and use the
eigenvalues equation.
To diagonalize a matrix, aim to find its eigenvalues and then write tr(A) = a + d = 1+2 = 3 and det(A) = ad-bc = 2-6 = 4.
them in a diagonal matrix. Recall that the square matrix Use the eigenvalues equation: λ²-tr(A λ + det(A) = 0.
size determines the degree of its characteristic polynomial Substitute values to get λ²-3λ-4 = 0 which is a quadratic
equation. For an instant, a 2x2 matrix has a quadratic characteristic equation that has two solutions: λ₁ = -1, λ₂ = 4.
polynomial equation that has two
08
Hence, the given matrix has two distinct eigenvalues. Matrix D is written as: Step 3: Write the diagonal matrix.
−1 0 0 0 0
D= and matrix A is diagonalizable since its eigenvalues are found. D = 0 −2 0
0 4
Non Diagonalizable Matrix A is diagonalizable since its eigenvalues are found. 0 0 4
Example 2:
2 6
A= Diagonalize the following matrix.
Example 1: 0 −1
−1 3 1 Solution:
A = 2 4 −2 Diagonalize the following matrix. Step 1: Find the matrix determinant and trace.
1 3 −1 det(A) = -2 and tr(A) = 1
Solution: Step 2: Write the characteristic polynomial equation and set it to zero,
then find its solutions
Step 1: Find the matrix determinant and trace. Since this is a 3x3 matrix then, λ²-tr(A)λ+det(A) = 0 and substitute: λ²-λ-2 = 0
the determinant is calculated as: which has two solutions: 2 and -1
det(A) = -1(-4+6)-3(-2+2)+1(6-4) = 0 and the trace is equal to: tr(A) = -1+4-1 = 2 Step 3: Write the diagonal matrix.
Step 2: Write the characteristic polynomial equation and set it to zero, then find 𝝀₁ 0 2 0
its solutions. As a 3x3 matrix, the characteristic polynomial equation is: D= and substitute to get: D =
0 𝝀₂ 0 −1
-λ³+tr(A)λ²-½(tr²(A)-tr(A))λ+det(A) = 0
To substitute in this formula, one more term is needed, the trace of the square Non Diagonalizable Matrix
of the matrix. So, A matrix is not diagonalizable when it does not have distinct
−1 3 1 −1 3 1 8 12 −8 eigenvalues. All matrices of the form nxm, non-square
A = 2 4 −2 2 4 −2 = 4 16 −4 matrices have no eigenvalues;
1 3 −1 1 3 −1 4 12 −4 hence, there are not diagonalizable. Solve the following example
and learn to recognize when is a matrix not diagonalizable.
which has the trace: tr(A²) = 8+16-4 = 20. Thus, 2 1 −5
-λ³+2λ²-½(4-20)λ+0 = -λ³+2λ²+8 λ = 0 For Example: H =( ) Find if the following matrix
0 −2 4
which has three solutions: 0, -2, and 4. So, the eigenvalues of matrix A in diagonalizable or not.
the example are 0, -2, and 4. Solution:
The first step is to check the matrix eigenvalues.
This matrix is not square, so one cannot find its eigenvalues. Thus, it is not
diagonalizable.
09
Application: Conclusion:

Matrices undergoing diagonalization play a pivotal role in diverse Diagonalization serves as a powerful mathematical tool, streamlining
applications, offering a streamlined approach to computations and unveiling matrix operations and uncovering inherent system structures.
fundamental characteristics of linear systems. The versatility of Its significance spans diverse fields, such as linear algebra, physics,
diagonalization is evident in several domains: computer science, and engineering, where it not only eases
computations but also unveils patterns and symmetry. In the realm
1,Solving Linear Systems: The process of tackling systems of linear equations of physics, diagonalization facilitates a deeper grasp of quantum
is simplified through diagonalization. When a matrix is diagonalizable, the mechanics, while in computer science, it finds application in techniques
solution to Ax = b becomes more straightforward by utilizing the diagonal like principal component analysis (PCA) for both dimensionality reduction
form D = = 𝑃−1 AP. and pattern recognition. The spectral theorem further underscores
2. Matrix Powers: Efficiency in computing matrix powers, 𝐴−1 , is enhanced the pivotal role of diagonalization in unravelling intricate mathematical
when dealing with diagonalizable matrices. The diagonal form facilitates structures.
straightforward exponentiation, providing valuable insights into the long-
term behaviour of dynamic systems.
3.Quantum Mechanics: Diagonalization finds application in quantum Reference:
mechanics, particularly in representing operators in a diagonal format. This
simplifies the study of quantum systems' properties and aligns with
principles such as the spectral theorem. 1)Higher Engineering Mathematics, 42nd Edition, Dr. B. S. Grewal,
4. Control Theory: In the realm of control theory, diagonalization is a Published by Khanna Publishers, June 2012
valuable tool for analysing and designing control systems. It simplifies the 2)www.study.com
comprehension of system dynamics and aids in the design of controllers for 3)Engineering Mathematics Volume I, B.K. Pal and K. Das, Published
stability and performance. by U. N. Dhur and Pvt. Ltd., August 2022
5. Structural Analysis: Diagonalization proves useful in structural engineering 4)https://www.geeksforgeeks.org/matrix-diagonalization/
for addressing eigenvalue problems. It assists in the analysis of stability and 5)https://math.okstate.edu/people/binegar/3013-S99/3013-116.pdf
vibrations within structures, with the diagonalized form aiding in the
identification of critical modes and frequencies.

10