Diagonalizable matrix

A matrix is diagonalizable if it is similar to a diagonal matrix — that is, if there exists an invertible matrix $P$ such that $P^{-1}AP = D$, where $D$ is diagonal. Equivalently, an n × n matrix is diagonalizable if and only if it has $n$ linearly independent eigenvectors, which occurs precisely when the geometric multiplicity of each eigenvalue equals its algebraic multiplicity. Diagonalizable matrices are the "nice" case of linear algebra: they can be fully understood by their eigenvalues and eigenvectors alone.

The diagonal form is extraordinarily useful. Powers of a diagonalizable matrix are trivial to compute: $A^k = PD^kP^{-1}$, where $D^k$ is just the diagonal entries raised to the $k$-th power. This makes diagonalization essential for solving linear recurrence relations, computing matrix exponentials $e^{At}$, and analyzing the long-term behavior of discrete and continuous dynamical systems. The matrix exponential, which solves the linear differential equation $\dot{x} = Ax$, becomes a sum of exponentials of eigenvalues when $A$ is diagonalizable.

Not all matrices are diagonalizable. A matrix fails to be diagonalizable when it is defective — when at least one eigenvalue has fewer linearly independent eigenvectors than its algebraic multiplicity. In this case, the matrix cannot be written as $PDP^{-1}$ for diagonal $D$, and one must resort to the Jordan normal form instead. The Jordan form generalizes diagonalization by allowing near-diagonal blocks, but at the cost of losing the simplicity and numerical stability of true diagonalization.

Diagonalizability is linear algebra's comfort zone — the assumption that every matrix can be reduced to independent one-dimensional actions. But the world is full of defective matrices: coupled oscillators with degenerate frequencies, Markov chains with transient states, and control systems with uncontrollable modes. The insistence on diagonalizability is not a mathematical necessity; it is a pedagogical convenience. We teach diagonalization first because it is clean, and we teach Jordan form as an afterthought because it is messy. This ordering is backwards. Defectiveness is not a pathology; it is the generic case in parameterized families of matrices, and it is where the interesting geometry lives.