The reduced Berggren dynamics is never ergodic.
ProvedPadicBerggren.B2_not_transitive_on_nullConeaether-cataloggeometry
The reduced Berggren dynamics is never ergodic. For every odd prime the hyperbolic
generator fails to be transitive on the p² − 1 nonzero null vectors mod p: no single orbit
can exhaust the null cone, because orbits have at most p + 1 elements while the punctured
null cone has p² − 1 = (p − 1)(p + 1) elements.
theorem PadicBerggren.B₂_not_transitive_on_nullCone(hp : p ≠ 2) (v : Fin 3 → ZMod p) :
¬ (∀ w ∈ (nullConeFinset p).erase 0, ∃ n : ℕ, (B₂ (ZMod p)) ^ n *ᵥ v = w) := by sorry
Formalization Note Transplanted verbatim from the Aether Catalog source Geometry/PadicBerggrenOrbits.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.
Upload correction: only the declared theorem identifier was normalized from PadicBerggren.B₂_not_transitive_on_nullCone to PadicBerggren.B2_not_transitive_on_nullCone; mathematical statement unchanged.
Preamble
-- Thm stub generated from Geometry/PadicBerggrenOrbits.lean import Mathlib import Definitions.Def_Geometry_PadicBerggrenDynamics import Definitions.Def_Geometry_PadicBerggrenNullCone /-! # Orbit structure of the reduced Berggren dynamics Building on `Catalog/Geometry/PadicBerggrenDynamics.lean` (the three Berggren generators as a dynamical system on `(ZMod (p^k))³`) and on `Catalog/Geometry/PadicBerggrenNullCone.lean` (the phase space has exactly `p²` points), this file settles the **orbit structure** of the generators on the null cone mod an odd prime. ## Main results * `PadicBerggren.pow_mod_eq` : an elementary periodicity reduction, `M^n = M^(n % m)` whenever `M^m = 1`. * `PadicBerggren.B₂_pow_le_p_add_one` : for every odd prime there is a period `m` with `1 ≤ m ≤ p + 1` and `B₂^m = 1` — either `p − 1` (split case, `2` a square mod `p`) or `p + 1` (inert case). So the *actual* period of the hyperbolic generator is at most `p + 1`, much smaller than the a priori bound `p² − 1`. * `PadicBerggren.B₂_orbit_card_le` : every `B₂`-orbit on `(ZMod p)³` has at most `p + 1` points. * `PadicBerggren.B₂_not_transitive_on_nullCone` : consequently, for every odd prime the hyperbolic generator is **never transitive** on the `p² − 1` nonzero null vectors: the reduced Berggren dynamics is *never ergodic* on the null cone, and there are at least `(p² − 1)/(p + 1) = p − 1` distinct orbits. This is the precise sense in which the p-adic picture differs from the real one, where the hyperbolic generator has dense orbits on the boundary. * `PadicBerggren.card_B₁_fixedPoints` : the unipotent generator fixes exactly `p` points — a whole isotropic line — whereas `B₂` fixes only the origin (`B₂_no_nonzero_fixed_point`). This is the counting form of the unipotent/hyperbolic spectral dichotomy. -/ open PadicBerggren open Matrix Finset variable (p : ℕ) [Fact p.Prime]
Formal statement
theorem PadicBerggren.B2_not_transitive_on_nullCone(hp : p ≠ 2) (v : Fin 3 → ZMod p) :
¬ (∀ w ∈ (nullConeFinset p).erase 0, ∃ n : ℕ, (B₂ (ZMod p)) ^ n *ᵥ v = w) := by sorrySource