Iterated function systems
An iterated function system describes a fractal with a finite collection of transformations—usually shifts, rotations, scales and shears—that repeatedly map space into itself.

Contracting maps and an attractor
If every transformation is contractive, the Hutchinson operator maps a set to the union of its transformed copies and has a unique compact fixed set. Repeating the operator from many starting shapes converges toward the same attractor.
For a finite family of contractions, the contraction mapping theorem—applied in a suitable metric on sets—guarantees a unique compact attractor. In practice, begin with any nonempty compact shape, map it through every function and unite the results. The sequence converges to the same limit. The final fern or carpet therefore does not depend on the starting cloud as long as the maps are genuinely contracting.
Deterministic and random rendering
A deterministic renderer applies every map to the current collection. The chaos game chooses maps randomly with suitable probabilities and plots the orbit after a warm-up. Under the right conditions both reveal the same attractor.
The deterministic algorithm draws every component copy at each stage and explains the construction clearly, but grows exponentially. The chaos game chooses one map per step and accumulates visited points. Selection probabilities mainly alter density and convergence speed rather than the geometric support, provided every map still occurs. For visually even sampling, probabilities should roughly track the area contributions of the component images.
A compact geometric language
IFS descriptions can encode ferns, trees, dusts and tiles in very little data. That economy inspired fractal compression, but finding transformations that faithfully represent an arbitrary photograph is a difficult inverse problem.
An IFS stores complex form in a few matrices, translations and weights. That compactness is both strength and limitation: it works beautifully when genuine affine repetitions exist and poorly for arbitrary photographs without matching subregions. Varying one matrix continuously changes the attractor in a traceable way. This makes an IFS a useful design space, not merely a recipe for familiar textbook fractals.
Sources and further reading
This article summarizes the following specialist sources in original wording. Accessed and editorially reviewed 12 August 2026.
- Iterated Function SystemWolfram MathWorld
- Fractal Geometry: Mathematical Foundations and ApplicationsWiley

