Bounded Koopman operator criterion Created 2026-09-28 Updated 2026-11-05
For a nonsingular transformation of a probability space, let be its Radon-Nikodym derivative. The Koopman operator extends to a bounded operator on exactly when , and then
Indeed, change of variables gives . Conversely, if the squared operator norm is at most , test to obtain for every measurable . The sets must therefore be null, proving almost everywhere and equality of the optimal constants. Nonsingularity supplies a well-defined composition on equivalence classes, while the essential bound supplies the norm estimate.
Koopman operator Created 2026-09-28 Updated 2026-11-05
For a measurable map , the Koopman operator acts on observables by composition:
It is linear in even if the state map is nonlinear. To act on measurable functions modulo equality almost everywhere, require to be a nonsingular transformation; otherwise changing on a null set could change on a positive-measure set. Boundedness on a chosen Lp space is an additional question, treated by the bounded Koopman operator criterion for .
For a measure-preserving transformation, change of variables gives , so the Koopman operator is an isometry. If has a measurable measure-preserving inverse modulo null sets, it is a unitary operator. The converse need not follow from isometry alone: for , the Fourier observable maps to , so the range misses odd Fourier modes and the operator is not onto.
For an irrational rotation of the circle by , . Thus observable eigenfunctions encode exact oscillatory correlations, a useful bridge between nonlinear dynamical systems and spectral theorem.
The map is a nonsingular transformation with respect to when
Equivalently, the pushforward measure satisfies .
For an essentially bounded observable , define the Koopman operator
Nonsingularity makes this well defined on almost-everywhere equivalence classes. By the Radon-Nikodym theorem,
Consequently the bounded Koopman operator criterion is