Taylor Series
Any smooth function can be approximated by a polynomial. Watch Taylor polynomials converge to their target function as you add more terms — a beautiful bridge between the algebraic and the transcendental.
How it works
A Taylor series represents a function as an infinite sum of terms calculated from the values of its derivatives at a single point. The Maclaurin series (Taylor series centered at 0) of f(x) is f(0) + f′(0)x + f″(0)x²/2! + f′′′(0)x³/3! + …. Each additional term captures one more degree of the function’s curvature at the center point.
Not all Taylor series converge everywhere. The series for ex converges for all x — remarkable, given that we are reconstructing an exponentially growing function from polynomials. But ln(1+x) converges only for −1 < x ≤ 1, and 1/(1−x) only for |x| < 1. The radius of convergence is determined by the nearest singularity in the complex plane, even if that singularity doesn’t appear on the real line. For 1/(1−x), the pole at x = 1 limits convergence. For ln(1+x), the branch point at x = −1 sets the boundary.
Taylor series are not merely theoretical constructs. Computers use them (or close relatives like Chebyshev polynomials and minimax approximations) to evaluate transcendental functions. When your calculator computes sin(0.5), it is essentially evaluating a carefully optimized polynomial approximation. The error of an n-term Taylor polynomial is bounded by the (n+1)-th term, giving a precise sense of how good the approximation is. This interplay between the infinite and the finite — capturing a transcendental function in a polynomial straitjacket — is one of the most elegant ideas in mathematics.