Definition and Core Idea
A power series is an infinite series of the form ∑_{n=0}^{∞} c_n (x − a)^n, where the coefficients c_n and the center a are fixed real or complex numbers and x is a variable. At its core, a power series combines infinitely many monomial terms scaled by constants, centered around a point a. Because each term is a power of (x − a) multiplied by a coefficient, the structure is simple and parametric, which makes power series both expressive and analyzable. They serve as local approximations to more complicated functions and, within their interval of convergence, behave like an infinite polynomial that can encode familiar operations term-by-term.
Convergence and Interval of Convergence
Because not infinite sums behave uniformly, it is essential to determine for which values of x a power series converges. The key tool is the radius of convergence R, a nonnegative number (possibly infinite) such that the series converges absolutely for |x − a| < R and diverges for |x − a| > R. At the boundary points x = a ± R, convergence must be checked separately, since the behavior can differ for each endpoint. Common techniques to find R include the ratio test, the root test, and comparison with known series. The set of x-values where the series converges forms the interval of convergence, which may be open, closed, or half-open, and it always has the symmetric structure (a − R, a + R), possibly including one or both endpoints.
Ratio Test for Radius
Given a power series ∑ c_n (x − a)^n with nonzero coefficients, define L = lim_{n→∞} |c_{n+1} / c_n| when the limit exists. If L is finite and positive, the radius of convergence is R = 1/L; if L = 0 then R = ∞, and if L is infinite then R = 0. This provides a practical computational method in many standard examples. When the limit does not exist, the root test using limsup_{n→∞} |c_n|^{1/n} offers a more general approach. These tools translate coefficient growth into precise geometric bounds on where the series converges.
Representing Functions and Taylor Series
A central use of power series is to represent functions as sums of simpler terms. When c_n = f^{(n)}(a) / n! for a smooth function f, the resulting power series is called the Taylor series of f centered at a. If a = 0, the series is known as a Maclaurin series. Within the radius of convergence, the Taylor series can equal the original function, enabling exact representations of exponentials, trigonometric functions, logarithms, and binomial expressions. Notably, differentiability and continuity properties of the original function are inherited termwise by the series inside its interval of convergence, allowing calculus operations to be performed on the series itself.
Common Standard Expansions
Several elementary functions have well-known power series that serve as building blocks. For example, the geometric series ∑_{n=0}^{∞} x^n equals 1/(1 − x) for |x| < 1 and provides a simple prototype. The exponential series ∑_{n=0}^{∞} x^n / n! converges for all real x, as do the sine and cosine series with alternating signs and odd/even powers. The natural logarithm and binomial series introduce rational and generalized binomial coefficients, respectively, illustrating how coefficient structure determines function behavior. These examples highlight how coefficient patterns directly control growth, symmetry, and convergence radius.
| Function | Power Series (centered at 0) | Interval of Convergence |
|---|---|---|
| 1 / (1 − x) | ∑_{n=0}^{∞} x^n | (−1, 1) |
| e^x | ∑_{n=0}^{∞} x^n / n! | (−∞, ∞) |
| sin x | ∑_{n=0}^{∞} (−1)^n x^{2n+1} / (2n+1)! | (−∞, ∞) |
| cos x | ∑_{n=0}^{∞} (−1)^n x^{2n} / (2n)! | (−∞, ∞) |
| ln(1 + x) | ∑_{n=1}^{∞} (−1)^{n+1} x^n / n | (−1, 1] |
Algebraic and Analytical Properties
Power series behave well under addition, subtraction, and multiplication, enabling formal manipulations similar to polynomials. Within the interval of convergence, a power series can be differentiated and integrated term by term, yielding new series with the same radius of convergence. This makes them suitable for solving differential equations, where solutions are expressed as series whose coefficients satisfy recurrence relations. Moreover, uniform convergence on compact subsets inside the radius of convergence justifies limit exchanges, allowing continuity, integrability, and differentiability to be studied through the coefficients and partial sums.
Applications in Analysis and Computation
In mathematical analysis, power series provide rigorous foundations for defining and manipulating transcendental functions. In numerical computation, truncated series (polynomial approximations) are used to evaluate functions efficiently when high precision is required and elementary function implementations are unavailable. Generating functions in combinatorics encode sequence data into power series, linking discrete counting problems with analytic tools. In physics and engineering, power series appear in perturbation methods, signal processing, and local linearization near equilibria. Across these domains, controlling approximation error through remainder estimates and understanding convergence behavior are essential for reliable use.
Practical Considerations and Limitations
Practical use of power series requires awareness of their limits. Outside the interval of convergence, the series diverges and cannot represent the function. At boundary points, convergence may or may not hold, so each case needs separate verification. For rapidly growing coefficients, the radius of convergence can be small, limiting the domain of usefulness. Numerical stability can degrade when evaluating high-degree terms directly, especially near the boundaries or for alternating series where partial sums oscillate. In such cases, transformations, scaling, or alternative representations (such as Padé approximants) may improve performance. Recognizing when a power series is the right tool—and when another method is preferable—is key to effective application.