The full-time exponent set equals the discrete return-time exponent set #
This module proves the set equality that lets the full-time special-flow Lyapunov-exponent set
be recognised as measurable. For a fixed value L : ℝ, the set of base points carrying the
full-time cover-cocycle growth rate L,
{x | Real.log ‖coverCocycle A T hτ hc hcpos (x, 0) t‖ / t → L as t → ∞},
coincides — pointwise in x, under a bounded roof c ≤ τ ≤ C — with the set of base points
carrying the discrete return-time growth rate L,
{x | Real.log ‖cocycle A (⇑T) n x‖ / returnTime T hτ n x → L as n → ∞}.
The two inclusions are the two halves of the between-returns squeeze, run per point rather than
μ-a.e.:
Full ⟹ discrete (
tendsto_returnTimeExp_of_coverCocycle). Sample the full-time limit along the return-time subsequencet = returnTime T hτ n x, which tends to+∞(returnTime_tendsto_atTop); the cover-cocycle norm at a return time equals the base cocycle norm (coverCocycle_returnTime_norm_eq), so the sampled full-time ratio is exactly the discrete return-time ratio. This direction needs only the positive lower roof bound.Discrete ⟹ full (
tendsto_coverCocycle_of_returnTimeExp). This is the per-point core ofcoverCocycle_tendsto_exponent_of_bddRoof: factor the full-time ratio as the time-distortion factorreturnTime (lapCount t x) x / t(which→ 1bylapCount_returnTime_div_tendsto_one, the bounded-roof input) times the discrete exponent ratio evaluated at the lap count (which→ Lby composing the discrete hypothesis withlapCount_tendsto_atTop). The product factorization islog_coverCocycle_div_eq_lapCount. This direction genuinely needs the uniform upper roof boundτ ≤ C.
The normal form of the discrete set is fixed exactly as
Real.log ‖cocycle A (⇑T) n x‖ / returnTime T hτ n x so that the downstream measurability rewrite
(Oseledets.Continuous.SuspensionExponentSetMeasurable) lands on a
MeasureTheory.measurableSet_tendsto shape with the per-index measurability supplied by
measurable_logNorm_div_returnTime.
This is the special-flow / flow-under-a-roof construction of Cornfeld–Fomin–Sinai, Ergodic Theory
(Springer 1982), Ch. 11 (special/suspension flows; Ambrose–Kakutani); the Lyapunov-exponent
analogue of Abramov's entropy formula h(flow) = h(base)/∫τ.
Main results #
Oseledets.tendsto_returnTimeExp_of_coverCocycle: full-time exponent ⟹ discrete return-time exponent (sampling along the returns).Oseledets.tendsto_coverCocycle_of_returnTimeExp: discrete return-time exponent ⟹ full-time exponent (the bounded-roof between-returns squeeze, per point).Oseledets.coverCocycle_exponent_set_eq: the two exponent sets coincide, for every valueL.
Full-time exponent ⟹ discrete return-time exponent. If the cover-cocycle log-norm rescaled
by the elapsed flow time t tends to L as the real t → ∞, then the discrete return-time ratio
Real.log ‖cocycle A (⇑T) n x‖ / returnTime T hτ n x tends to L as n → ∞. The discrete ratio is
the full-time ratio sampled along the return-time subsequence t = returnTime T hτ n x (which tends
to +∞, returnTime_tendsto_atTop), and the cover-cocycle norm at a return time equals the base
cocycle norm (coverCocycle_returnTime_norm_eq). Needs only the lower roof bound 0 < c ≤ τ.
Discrete return-time exponent ⟹ full-time exponent. If the discrete return-time ratio
Real.log ‖cocycle A (⇑T) n x‖ / returnTime T hτ n x tends to L as n → ∞, then under a bounded
roof c ≤ τ ≤ C the full-time cover-cocycle log-norm rescaled by t tends to L as the real
t → ∞. This is the per-point core of coverCocycle_tendsto_exponent_of_bddRoof: the full-time
ratio factors (log_coverCocycle_div_eq_lapCount) as the time-distortion factor
returnTime (lapCount t x) x / t (→ 1 by lapCount_returnTime_div_tendsto_one) times the
discrete exponent ratio sampled at the lap count (→ L by composing the hypothesis with
lapCount_tendsto_atTop). The product of the two limits is L · 1 = L. Needs the upper roof bound
τ ≤ C.
The full-time exponent set equals the discrete return-time exponent set. For a fixed value
L, under a bounded roof c ≤ τ ≤ C, the set of base points whose full-time cover-cocycle growth
rate is L coincides with the set of base points whose discrete return-time growth rate is L:
{x | log ‖coverCocycle (x, 0) t‖ / t → L} = {x | log ‖cocycle A (⇑T) n x‖ / returnTime n x → L}.
This is the two-sided between-returns squeeze, run per point. The forward inclusion samples along
the returns (tendsto_returnTimeExp_of_coverCocycle), the reverse runs the bounded-roof squeeze
(tendsto_coverCocycle_of_returnTimeExp).