Diagonalization

Definition

Diagonalization is the process of decomposing a square matrix $A$ into a product of three matrices involving its eigenvalues and eigenvectors: $A = X \Lambda X^{-1}$ where $X$ is a matrix whose columns are the linearly independent eigenvectors of $A$, and $\Lambda$ (Lambda) is a diagonal matrix containing the corresponding eigenvalues.

• How to read: “The matrix A equals X times Lambda times X inverse.”
• Meaning: In the eigenvector basis, $A$ acts as simple scaling by eigenvalues on the diagonal of $\Lambda$.

Why It Matters

In a “tangled” matrix, variables interfere with each other, making long-term prediction impossible. Diagonalization is the ultimate “untangling” operation, isolating the core growth rates (eigenvalues) from the mess. Without it, calculating the billionth step of a dynamical system or the state of a quantum atom would be computationally impossible.

Core Concepts

• Diagonalizability Condition: A matrix $A$ of size $n \times n$ is diagonalizable if and only if it has $n$ linearly independent eigenvectors.
• The Diagonal Matrix $\Lambda$: $\Lambda = \begin{bmatrix} \lambda_1 & 0 & \dots & 0 \\ 0 & \lambda_2 & \dots & 0 \\ \vdots & \vdots & \ddots & \vdots \\ 0 & 0 & \dots & \lambda_n \end{bmatrix}$
  • How to read: “The matrix Lambda equals the diagonal matrix with lambda one through lambda n on the diagonal and zeros elsewhere.”
  • Meaning: Each eigenvalue $\lambda_i$ sits on the diagonal; off-diagonal entries are zero because eigenvectors decouple the action of $A$.
• Multiplicity Rule: For a matrix to be diagonalizable, the Geometric Multiplicity (number of independent eigenvectors) must equal the Algebraic Multiplicity (number of times the eigenvalue appears as a root) for every eigenvalue.
• Powers of a Matrix: Diagonalization allows for efficient calculation of matrix powers: $A^k = X \Lambda^k X^{-1}$
  • How to read: “The matrix A to the k equals X times the quantity Lambda to the k times X inverse.”
  • Meaning: Once diagonalized, repeated application of $A$ is just repeated scaling by the eigenvalues — raise each diagonal entry of $\Lambda$ to the power $k$ instead of multiplying $A$ by itself $k$ times. Since $\Lambda$ is diagonal, $\Lambda^k$ is simply the diagonal elements raised to the power of $k$.

Connected Concepts

• Eigenvalues and Eigenvectors
• Linear Transformation
• Change Of Basis
• The Spectral Theorem