Power Series Convergence

Visualize how the sequence of partial sums behaves inside and outside the radius of convergence.

Power Series Convergence

Concept Overview

A power series is an infinite series of the form Σ c_n(x − a)ⁿ. It represents a function as an infinite polynomial, but this representation is only valid where the infinite sum evaluates to a finite number — that is, where the series converges. The interval of convergence defines the domain over which the power series accurately models the target function. Understanding convergence is essential for using power series reliably in analysis, physics, and computation.

Mathematical Definition

A power series centered at x = a has the general form:

f(x) = Σ (n=0 to ∞) c_n(x − a)ⁿ = c₀ + c₁(x − a) + c₂(x − a)² + ···

When a = 0, this simplifies to a Maclaurin series:

f(x) = Σ (n=0 to ∞) c_nxⁿ

The partial sum S_N(x) is the sum of the first N terms. The series converges if the sequence of partial sums has a finite limit as N → ∞.

Radius of Convergence (R)

Every power series has a characteristic radius of convergence R. There are exactly three possibilities for any power series centered at a:

The series converges only at x = a. (R = 0)
The series converges absolutely for all real numbers. (R = ∞)
There exists a positive number R such that the series converges absolutely when |x − a| < R and diverges when |x − a| > R.

The most common method to find R is the Ratio Test. For a series Σ a_n, evaluate:

R = lim (n→∞) | a_n / a_n+1 |

The behavior at the exact boundary points (|x − a| = R) must be tested individually using specialized tests such as the Alternating Series Test or the p-series test.

Key Examples

The Geometric Series is the foundation for analyzing many other power series:

1 / (1 − x) = Σ (n=0 to ∞) xⁿ = 1 + x + x² + x³ + ···

This series has R = 1. For x = 2, the partial sums grow without bound (1 + 2 + 4 + 8 + ···), diverging rapidly from the true function value of −1.

The Exponential Series has a Maclaurin series that converges everywhere (R = ∞):

e^x = Σ (n=0 to ∞) xⁿ / n! = 1 + x + x²/2! + x³/3! + ···

Because n! grows faster than any power xⁿ, the terms shrink to zero regardless of how large x is.

Key Concepts

Absolute convergence: A series Σ a_n converges absolutely if Σ |a_n| converges. Inside the radius of convergence, a power series always converges absolutely.
Interval of convergence: The set of all x for which the series converges, typically the open interval (a − R, a + R) with boundary points checked separately.
Truncation error: The difference between the true function value and the partial sum S_N. For a degree-N approximation, the error is on the order of (x − a)^N+1, bounded by the Lagrange remainder term.
Alternating series: When x is negative in the geometric series, successive partial sums alternate above and below the true value, which makes the error easy to bound.
Analytic functions: A function is analytic at a point if its power series converges to the function in some neighborhood of that point. Most functions in physics and engineering are analytic on their domains.

Historical Context

The systematic study of power series began in the 17th century. Isaac Newton used infinite series to represent functions such as (1 + x)ⁿ for non-integer n as early as 1665, laying groundwork for the general theory. Brook Taylor (1685–1731) formalized the expansion of functions as polynomial series in his 1715 work Methodus Incrementorum.

A rigorous theory of convergence came later through the work of Augustin-Louis Cauchy (1789–1857) and Karl Weierstrass (1815–1897), who established the definitions of limit, continuity, and uniform convergence that underpin modern analysis. The concept of the radius of convergence is attributed to Cauchy, who derived it using the root and ratio tests.

Real-world Applications

Differential equations: Power series methods solve equations (such as Bessel's equation) where elementary closed-form solutions do not exist.
Computer science: Hardware and software evaluation of transcendental functions (sin, cos, ln) relies on polynomial approximations derived from convergent power series.
Physics: Perturbation theory expresses physical quantities as power series of a small parameter (e.g., coupling strength in quantum mechanics). Convergence of these series determines whether the approximation is physically meaningful.
Probability: Generating functions in statistics, which encode probability distributions, are fundamentally power series.

Related Concepts

Taylor Series — specific power series generated by the derivatives of a function
Fourier Series — representing periodic functions as sums of sinusoids
Analytic Continuation — extending the domain of a function defined by a power series beyond its original radius of convergence
Numerical Methods — finite difference schemes derived from Taylor expansions

Power Series Convergence