Root of a Function (Zero of a Function)

Terminology and multiplicity

When a root x0 is such that f(x0)=0 but f'(x0)≠0 (for differentiable f), it is called a simple root. If f and several of its derivatives vanish at x0, the root has higher multiplicity: a root of multiplicity m means the factor (x−x0)^m divides the function in a local or algebraic sense. Multiplicity affects both algebraic counts and numerical behaviour: many iterative methods converge more slowly near multiple roots.

Polynomials and complex roots

Polynomials are a central case because of their algebraic structure; see polynomials for background. The fundamental theorem of algebra, discussed in standard texts at this reference, states that every nonconstant polynomial with complex coefficients has at least one complex root. Combining that result with division lemmas such as the polynomial remainder theorem shows that a nonzero polynomial of degree n has exactly n complex roots when counted with multiplicity. For polynomials with real coefficients, nonreal roots occur in conjugate pairs.

Existence results for general functions

Continuity and differentiability give criteria for the existence and location of real roots. The intermediate value theorem guarantees at least one real root in any interval where a continuous function changes sign. Analytic functions have isolated zeros unless the function is identically zero on a region, and methods from complex analysis give powerful tools to count and locate zeros in the complex plane.

Methods for finding roots

Exact algebraic solutions exist only in special cases (linear, quadratic, cubic and quartic formulae under certain forms). In practice, numerical methods are ubiquitous. Robust bracketing methods such as bisection locate a root when a sign change is known; open methods such as Newton–Raphson have faster local convergence but require a good initial guess and derivative information. Variants include the secant method, regula falsi, and higher-order schemes. For polynomials, dedicated algorithms (for example companion-matrix approaches or simultaneous root methods) compute all roots; iterative approaches may use deflation to remove found factors carefully to avoid numerical instability.

• Bracketing (bisection): reliable if a sign change is known.
• Newton-type methods: fast near simple roots, slower or unstable near multiple roots.
• Secant and hybrid methods: tradeoffs between robustness and speed.
• Special polynomial solvers: eigenvalue-based or simultaneous root iterations for all complex roots.

When implementing algorithms one must consider finite precision, stopping criteria, and multiplicity detection. Software documentation and numerical libraries provide practical guidance; see introductory materials and numerical references at root-finding resources and general polynomial guides.

Applications and related concepts

Roots appear throughout mathematics and applications: equilibrium points in physics, eigenvalue problems in linear algebra, boundary conditions in differential equations, and optimality conditions in optimization. Related concepts include poles of meromorphic functions (where values blow up rather than vanish), extrema (where derivatives vanish), and solution sets of systems of equations, which generalize root-finding to vector-valued functions and require tools from linear algebra and numerical continuation.

For further reading consult elementary treatments of equations and functions, numerical analysis introductions to root-finding, and algebraic or complex-analysis sources that elaborate on existence theorems and multiplicity; see references on the fundamental theorem at this source, division facts like the remainder theorem, and degree-related discussions at degree n. Practical software notes and examples are available at common computational guides and library documentation referenced at root-finding resources and in general polynomial material.