Circle packing theorem

The circle packing theorem (also called the Koebe–Andreev–Thurston theorem) states that for every connected planar graph, there exists a circle packing in the plane — a configuration of circles whose tangency relations match the graph's adjacency structure — and this packing is unique up to Möbius transformations. The theorem bridges discrete geometry and complex analysis: it provides a discrete analogue of the Riemann mapping theorem, replacing smooth conformal maps with combinatorial patterns of touching circles.

The theorem's power lies in its constructiveness. Where the Riemann mapping theorem guarantees existence without providing a method, circle packings can be computed by iterative algorithms that converge to the unique packing. This makes them a practical tool for numerical conformal mapping: given a polygonal domain, one can construct a circle packing that approximates the conformal map to the disk, with error bounds that improve as the packing refines.

Circle packings have found applications in medical imaging (flattening brain surfaces), computer graphics (texture mapping), and the study of random planar graphs. They also provide a discrete framework for understanding Teichmüller theory and the geometry of surfaces.

The circle packing theorem is a reminder that continuity is not the only path to rigor. The discrete can approximate the continuous, the combinatorial can capture the analytic, and a theorem about touching circles can say as much about conformal structure as any integral formula. The prejudice that discrete mathematics is somehow less profound than analysis is exactly that — a prejudice, and one that the circle packing theorem refutes with geometric clarity.