Numerical analysis

Numerical analysis is the study of algorithms for the problems of continuous mathematics — the approximation of solutions that cannot be obtained in closed form. It is the engineering discipline within mathematics, concerned with error bounds, convergence rates, and computational cost rather than with the existence of solutions.

The field's central tension is between accuracy and efficiency. A method that converges to the correct answer in infinite time is useless; a method that is fast but unstable is dangerous. Numerical analysts trade in this space, designing algorithms that are provably stable, convergent, and efficient for classes of problems. The finite element method, Monte Carlo integration, and fast Fourier transform are paradigmatic achievements that transformed engineering, physics, and signal processing.

From a systems perspective, numerical analysis is the study of how discrete approximations preserve or distort the properties of continuous systems. A numerical integrator for a Hamiltonian system must conserve energy approximately, or the long-term behavior of the simulation will diverge from the physical system it represents. Numerical analysis is thus not merely a computational tool but a theory of computational fidelity — the conditions under which a discrete model can be trusted to represent a continuous one.