Characteristic equation: det(A - λI) = 0, where I is the identity matrix
- Form the characteristic matrix A - λI
- Calculate the determinant: det(A - λI)
- Solve the characteristic equation: det(A - λI) = 0
A - λI = \(\begin{bmatrix} 4 & -2 \\ 1 & 1 \end{bmatrix} - λ\begin{bmatrix} 1 & 0 \\ 0 & 1 \end{bmatrix}\)
= \(\begin{bmatrix} 4-λ & -2 \\ 1 & 1-λ \end{bmatrix}\)
det(A - λI) = (4-λ)(1-λ) - (-2)(1)
= (4-λ)(1-λ) + 2
= 4 - 4λ - λ + λ² + 2
= λ² - 5λ + 6
λ² - 5λ + 6 = 0
λ² - 5λ + 6 = 0
Factor: (λ - 3)(λ - 2) = 0
So: λ₁ = 3, λ₂ = 2
Trace(A) = 4 + 1 = 5 = λ₁ + λ₂ = 3 + 2 ✓
det(A) = (4)(1) - (-2)(1) = 4 + 2 = 6 = λ₁ × λ₂ = 3 × 2 ✓
Eigenvalues: λ₁ = 3, λ₂ = 2
The characteristic equation is λ² - 5λ + 6 = 0 with eigenvalues λ₁ = 3 and λ₂ = 2
• Characteristic equation: det(A - λI) = 0
• Trace property: Sum of eigenvalues equals trace
• Determinant property: Product of eigenvalues equals determinant
Diagonalization: A = PDP⁻¹ where D is diagonal matrix of eigenvalues and P has eigenvectors as columns
Solve (A - 3I)v = 0
\(\begin{bmatrix} 4-3 & -2 \\ 1 & 1-3 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
\(\begin{bmatrix} 1 & -2 \\ 1 & -2 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
This gives: x - 2y = 0, so x = 2y
Choose y = 1, so x = 2: v₁ = \(\begin{bmatrix} 2 \\ 1 \end{bmatrix}\)
Solve (A - 2I)v = 0
\(\begin{bmatrix} 4-2 & -2 \\ 1 & 1-2 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
\(\begin{bmatrix} 2 & -2 \\ 1 & -1 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
This gives: 2x - 2y = 0, so x = y
Choose x = 1, so y = 1: v₂ = \(\begin{bmatrix} 1 \\ 1 \end{bmatrix}\)
P = [v₁ v₂] = \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix}\)
D = \(\begin{bmatrix} 3 & 0 \\ 0 & 2 \end{bmatrix}\) (diagonal matrix with eigenvalues)
For P = \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix}\):
det(P) = (2)(1) - (1)(1) = 1
P⁻¹ = \(\frac{1}{1}\begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix} = \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
PDP⁻¹ = \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix} \begin{bmatrix} 3 & 0 \\ 0 & 2 \end{bmatrix} \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
= \(\begin{bmatrix} 6 & 2 \\ 3 & 2 \end{bmatrix} \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
= \(\begin{bmatrix} 6-2 & -6+4 \\ 3-2 & -3+4 \end{bmatrix} = \begin{bmatrix} 4 & -2 \\ 1 & 1 \end{bmatrix} = A\) ✓
A = PDP⁻¹ where P = \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix}\) and D = \(\begin{bmatrix} 3 & 0 \\ 0 & 2 \end{bmatrix}\)
• Diagonalization: A = PDP⁻¹
• Eigenvector matrix: P has eigenvectors as columns
• Diagonal matrix: D has eigenvalues on diagonal
Diagonalizable matrix: Matrix with n linearly independent eigenvectors for n×n matrix
A - λI = \(\begin{bmatrix} 2-λ & 1 \\ 0 & 2-λ \end{bmatrix}\)
det(A - λI) = (2-λ)² = 0
So λ = 2 with algebraic multiplicity 2
(A - 2I)v = 0
\(\begin{bmatrix} 0 & 1 \\ 0 & 0 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
This gives: y = 0, and x can be any value
So all eigenvectors are of the form \(\begin{bmatrix} x \\ 0 \end{bmatrix} = x\begin{bmatrix} 1 \\ 0 \end{bmatrix}\)
There is only one linearly independent eigenvector: \(\begin{bmatrix} 1 \\ 0 \end{bmatrix}\)
For a 2×2 matrix to be diagonalizable, we need 2 linearly independent eigenvectors
The geometric multiplicity of λ = 2 is 1 (dimension of eigenspace)
Since geometric multiplicity (1) ≠ algebraic multiplicity (2), the matrix is not diagonalizable
This matrix represents a shear transformation combined with scaling
It has only one direction (the x-axis) that is preserved under the transformation
Such matrices are called defective matrices
Matrix is not diagonalizable
Matrix A is not diagonalizable because it has only one linearly independent eigenvector
• Diagonalizability condition: n linearly independent eigenvectors for n×n matrix
• Algebraic vs geometric multiplicity: Must be equal for diagonalizability
• Defective matrix: Has fewer eigenvectors than algebraic multiplicity
Matrix powers: If A = PDP⁻¹, then Aⁿ = PDⁿP⁻¹ where Dⁿ has eigenvalues raised to nth power
A = PDP⁻¹ where:
P = \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix}\), D = \(\begin{bmatrix} 3 & 0 \\ 0 & 2 \end{bmatrix}\), P⁻¹ = \(\begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
For a diagonal matrix, raising to a power means raising each diagonal element to that power:
D⁵ = \(\begin{bmatrix} 3^5 & 0 \\ 0 & 2^5 \end{bmatrix} = \begin{bmatrix} 243 & 0 \\ 0 & 32 \end{bmatrix}\)
A⁵ = \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix} \begin{bmatrix} 243 & 0 \\ 0 & 32 \end{bmatrix} \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
= \(\begin{bmatrix} 2 & 1 \\ 1 & 1 \end{bmatrix} \begin{bmatrix} 243 & 0 \\ 0 & 32 \end{bmatrix} \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
= \(\begin{bmatrix} 486 & 32 \\ 243 & 32 \end{bmatrix} \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}\)
= \(\begin{bmatrix} 486-32 & -486+64 \\ 243-32 & -243+64 \end{bmatrix}\)
= \(\begin{bmatrix} 454 & -422 \\ 211 & -179 \end{bmatrix}\)
We could compute A⁵ directly, but that would require 4 matrix multiplications
Using diagonalization, we only needed to raise scalars to the 5th power
For any power Aⁿ: Aⁿ = P\(\begin{bmatrix} 3^n & 0 \\ 0 & 2^n \end{bmatrix}\)P⁻¹
This shows the computational advantage of diagonalization
A⁵ = \(\begin{bmatrix} 454 & -422 \\ 211 & -179 \end{bmatrix}\)
• Matrix powers: Aⁿ = PDⁿP⁻¹ when A = PDP⁻¹
• Diagonal matrix powers: Raise diagonal elements to power
• Computational efficiency: Diagonalization greatly simplifies power calculations
Symmetric matrix: Matrix where Aᵀ = A, always has real eigenvalues and orthogonal eigenvectors
S - λI = \(\begin{bmatrix} 3-λ & 1 \\ 1 & 3-λ \end{bmatrix}\)
det(S - λI) = (3-λ)² - 1 = 9 - 6λ + λ² - 1 = λ² - 6λ + 8
λ² - 6λ + 8 = 0
(λ - 4)(λ - 2) = 0
So λ₁ = 4, λ₂ = 2
(S - 4I)v = 0
\(\begin{bmatrix} -1 & 1 \\ 1 & -1 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
This gives: -x + y = 0, so y = x
Choose x = 1: v₁ = \(\begin{bmatrix} 1 \\ 1 \end{bmatrix}\)
(S - 2I)v = 0
\(\begin{bmatrix} 1 & 1 \\ 1 & 1 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}\)
This gives: x + y = 0, so y = -x
Choose x = 1: v₂ = \(\begin{bmatrix} 1 \\ -1 \end{bmatrix}\)
v₁ · v₂ = (1)(1) + (1)(-1) = 1 - 1 = 0
Since the dot product is zero, the eigenvectors are orthogonal ✓
P = [v₁ v₂] = \(\begin{bmatrix} 1 & 1 \\ 1 & -1 \end{bmatrix}\)
D = \(\begin{bmatrix} 4 & 0 \\ 0 & 2 \end{bmatrix}\)
Since eigenvectors are orthogonal, we can normalize them to get an orthogonal matrix
P = \(\begin{bmatrix} 1/\sqrt{2} & 1/\sqrt{2} \\ 1/\sqrt{2} & -1/\sqrt{2} \end{bmatrix}\) (orthogonal matrix)
For symmetric matrices: A = PDPᵀ where P is orthogonal
For symmetric matrix S, eigenvectors are orthogonal and S = PDPᵀ
• Symmetric matrix property: Always diagonalizable with orthogonal eigenvectors
• Orthogonal diagonalization: A = PDPᵀ for symmetric A
• Real eigenvalues: Symmetric matrices have real eigenvalues
Characteristic equation: det(A - λI) = 0, whose roots are eigenvalues of A
Diagonalizable matrix: Matrix A such that A = PDP⁻¹ where D is diagonal
Algebraic multiplicity: Number of times an eigenvalue appears as a root of characteristic equation
Geometric multiplicity: Dimension of eigenspace for an eigenvalue
- Find characteristic equation: det(A - λI) = 0
- Solve for eigenvalues: Find roots of characteristic polynomial
- Find eigenvectors: For each eigenvalue λ, solve (A - λI)v = 0
- Check diagonalizability: Verify n linearly independent eigenvectors for n×n matrix
- Form diagonalization: A = PDP⁻¹
• Characteristic equation: det(A - λI) = 0
• Trace property: Sum of eigenvalues = trace(A)
• Determinant property: Product of eigenvalues = det(A)
• Diagonalizability: n linearly independent eigenvectors for n×n matrix
• Algebraic vs geometric: geom ≤ alg for each eigenvalue
• Symmetric matrices: Always diagonalizable with orthogonal eigenvectors
• Matrix powers: Aⁿ = PDⁿP⁻¹ when A = PDP⁻¹
• Characteristic polynomial roots
• Diagonalizability conditions
• Matrix power computations
Analysis: The visualization shows how different matrix properties affect diagonalization.
- Real eigenvalues are necessary for diagonalization over reals
- Independent eigenvectors are sufficient for diagonalization
- Symmetric matrices always diagonalizable