Logistic map

The logistic map is a discrete-time dynamical system defined by x_n+1 = rx_n(1 − x_n), usually with 0 ≤ x_n ≤ 1 and 0 ≤ r ≤ 4.¹ The variable can represent a normalized population, while r combines reproductive potential and the strength of density-dependent limitation. The factor x_n produces growth proportional to the existing population; the factor 1 − x_n suppresses growth as the population approaches the carrying scale.

For 0 < r < 1, every orbit tends toward zero. Between 1 and 3, a nonzero fixed point, x* = 1 − 1/r, attracts typical initial conditions. The map is important not because this biological interpretation is universally realistic, but because the same compact equation exposes how nonlinear feedback can generate complicated long-term behavior.

As r increases beyond 3, the stable fixed point loses stability through a period-doubling bifurcation, and stable cycles of periods 2, 4, 8, and higher appear in succession.¹ The bifurcation values accumulate near r ≈ 3.569945672, after which chaotic bands occur, although periodic windows remain embedded within them.

This structure is often displayed with a bifurcation diagram: after discarding transients, many iterates are plotted against r. The resulting branching pattern shows both the orderly cascade and the interleaving of periodic and chaotic regimes. The period-three theorem of Li and Yorke established a striking general principle for continuous interval maps: the existence of a period-three orbit entails orbits of every positive integer period and chaotic behavior in a precise mathematical sense.²

The logistic map became a central example of universality: unrelated one-dimensional maps with a smooth single-humped shape can share the same asymptotic ratios in their period-doubling cascades.³ The parameter intervals between successive doublings shrink by a factor approaching the Feigenbaum constant, δ ≈ 4.669201609. This number is not specific to the logistic polynomial; it reflects a broader universality class.

Chaotic orbits are deterministic but highly sensitive to initial conditions. Nearby starting values can separate exponentially for a time, a feature quantified by a positive Lyapunov exponent, before the bounded state space and nonlinear folding limit that separation. Numerical work must distinguish genuine asymptotic behavior from finite-precision artifacts: computer iterates eventually repeat because floating-point arithmetic represents only finitely many states. The map therefore serves both as a theoretical model and as a test case for numerical dynamical-systems methods.⁴

The logistic map has an exactly solvable edge case at r = 4. Under the substitution x = sin²(πθ), its iteration is related to angle doubling, which explains why its chaotic behavior can be analyzed through an expanding map rather than treated as mere visual irregularity. At this parameter, typical long-run states follow the invariant density 1/[π√(x(1 − x))], which concentrates near the endpoints rather than distributing values uniformly.

The model also illustrates why “chaos” does not mean that every parameter value is chaotic. Stable periodic windows, including a prominent period-three window, occur inside the chaotic region. In applications, estimated parameters and observations are finite and noisy, so apparent irregularity cannot by itself establish deterministic chaos. The logistic equation’s historical value lies in making these distinctions visible with a single recurrence: equilibrium, oscillation, bifurcation, universality, and chaos all arise within one bounded polynomial family.¹

Bifurcation

A qualitative change in a system’s long-term behavior caused by varying a parameter.

Fixed point

A value x* satisfying f(x*) = x*, so iteration leaves it unchanged.

Lyapunov exponent

A measure of the average exponential rate at which nearby trajectories separate or converge.

Period-doubling cascade

A sequence in which a stable cycle successively doubles its period as a control parameter changes.