Interval exchange transformation
From Wikipedia, the free encyclopedia
In mathematics, an interval exchange transformation is a kind of dynamical system that generalises the idea of a circle rotation. The phase space consists of the unit interval, and the transformation acts by cutting the interval into several subintervals, and then permuting these subintervals.
[edit] Formal Definition
Let n > 0 and let π be a permutation on . Consider a vector of positive real numbers (the widths of the subintervals), satisfying . We define a map called the interval exchange transformation associated to the pair (π,λ) as follows. For let and let . Then for , define Tπ,λ(x) = x − ai + a'i if x lies in the subinterval [ai,ai + λi). Thus Tπ,λ acts on each subinterval of the form [ai,ai + λi) by an orientation-preserving isometry, and it rearranges these subintervals so that the subinterval at position i is moved to position π(i).
[edit] Properties
Any interval exchange transformation Tπ,λ is a bijection of [0,1] to itself which preserves Lebesgue measure. It is not usually continuous at each point ai (but this depends on the permutation π).
The inverse of the interval exchange transformation Tπ,λ is again an interval exchange transformation. In fact, it is the transformation where for all .
If n = 2 and π = (12) (in cycle notation), and if we join up the ends of the interval to make a circle, then Tπ,λ is just a circle rotation. The Weyl equidistribution theorem then asserts that if the length λ1 is irrational, then Tπ,λ is uniquely ergodic. Roughly speaking, this means that the orbits of points of [0,1] are uniformly evenly distributed. On the other hand, if λ1 is rational then each point of the interval is periodic, and the period is the denominator of λ1 (written in lowest terms).
If n > 2, and provided π satisfies certain non-degeneracy conditions, a deep theorem due independently to W.Veech and to H.Masur asserts that for almost all choices of λ in the unit simplex the interval exchange transformation Tπ,λ is again uniquely ergodic. However, for there also exist choices of (π,λ) so that Tπ,λ is ergodic but not uniquely ergodic. Even in these cases, the number of ergodic invariant measures of Tπ,λ is finite, and is at most n.