The Diagonalization Theorem
You'll learn: the precise condition that determines when a square matrix can be diagonalized.
The Core Idea
We've seen what diagonalization is—expressing a matrix A as PDP⁻¹ where D is diagonal. But when is this possible?
The Diagonalization Theorem gives us the answer: An n×n matrix A is diagonalizable if and only if it has n linearly independent eigenvectors.
Think of it like this: to diagonalize A, we need to build the matrix P whose columns are eigenvectors. For P to be invertible (so we can write A = PDP⁻¹), its columns must be linearly independent. Since P is n×n, we need exactly n such eigenvectors.
Why This Matters
This theorem transforms diagonalization from a vague hope into a checkable condition:
- Find all eigenvalues of the n×n matrix
- Find the eigenvectors for each eigenvalue (build the eigenspaces)
- Count them: if you can gather n linearly independent eigenvectors total across all eigenspaces, you're done—the matrix is diagonalizable
If any eigenvalue has geometric multiplicity less than its algebraic multiplicity (fewer independent eigenvectors than the multiplicity of the root), you won't reach n independent eigenvectors, and diagonalization fails.
Example Check
For a 3×3 matrix, you need exactly 3 linearly independent eigenvectors. If one eigenvalue has algebraic multiplicity 2 but only produces 1 independent eigenvector, you can only get 2 total—not enough!
Key Takeaway: A matrix is diagonalizable ↔ it has n linearly independent eigenvectors, giving you the invertible matrix P needed for A = PDP⁻¹.