Lemma 1 — DKF equation and asymptotic concentration
ProvedDKFAsymptotics.lemma1Fix a positive finite state dimension and a stationary linear Gaussian model with transition density and stationary density , where are symmetric positive definite and .
For sequences of Gaussian densities and with positive-definite covariances, define
At every index where the precision matrix is invertible and is positive definite, the normalized density update is well defined, is a probability density, and equals at every .
Furthermore, for every probability measure and state , if and , then the update is eventually valid with positive-definite covariance, is eventually a Gaussian probability density, and converges weakly to . The measure may be degenerate; no invertibility of is assumed.
import Definitions.Def_DKFAsymptotics_Assumptions noncomputable section open MeasureTheory Filter Matrix open scoped Topology ENNReal
namespace DKFAsymptotics
theorem lemma1 {d : ℕ} [NeZero d] (M : Model d)
(a b' : ℕ+ → State d) (V U : ℕ+ → Covariance d)
(hV : ∀ n, (V n).PosDef) (hU : ∀ n, (U n).PosDef)
(P : ProbabilityMeasure (State d)) (b : State d) :
(∀ n, ValidStep M ⟨a n, V n⟩ ⟨b' n, U n⟩ →
UpdateDefined M (gaussianPDF (a n) (V n)) (gaussianPDF (b' n) (U n)) ∧
(∀ z, updateDensity M (gaussianPDF (a n) (V n)) (gaussianPDF (b' n) (U n)) z =
gaussianPDF (dkfStep M ⟨a n, V n⟩ ⟨b' n, U n⟩).mean
(dkfStep M ⟨a n, V n⟩ ⟨b' n, U n⟩).covariance z) ∧
IsPDF (updateDensity M (gaussianPDF (a n) (V n)) (gaussianPDF (b' n) (U n)))) ∧
(Weakly (fun n => gaussianPDF (a n) (V n)) P →
Weakly (fun n => gaussianPDF (b' n) (U n)) (Measure.dirac b) →
(∀ᶠ n in atTop,
ValidStep M ⟨a n, V n⟩ ⟨b' n, U n⟩ ∧
UpdateDefined M (gaussianPDF (a n) (V n)) (gaussianPDF (b' n) (U n)) ∧
IsGaussianPDF (updateDensity M (gaussianPDF (a n) (V n))
(gaussianPDF (b' n) (U n)))) ∧
Weakly (fun n => updateDensity M (gaussianPDF (a n) (V n))
(gaussianPDF (b' n) (U n))) (Measure.dirac b)) := by sorry
end DKFAsymptoticsRead-back
What the Lean code literally says, in plain math · GPT-6 (Codex independent sub-agent)
Read-back of DKFAsymptotics.lemma1
For every positive integer dimension (so is excluded), identify the state space with , with its usual topology and product Lebesgue measure . Let be real matrices such that and are symmetric positive definite and . Here symmetric positive definite means and for every nonzero . Let and be arbitrary sequences of vectors in , and let and be sequences of real matrices, with and symmetric positive definite for every positive integer . Let be any probability measure on and let be any vector. Define, for a vector , a real matrix , and ,
For every positive integer , put
The integral defining is the extended nonnegative Lebesgue integral, and is the usual real value when , with . All these formulas are total: the inverse of a singular matrix is the zero matrix, real division by zero yields zero, and the real square root of a nonpositive number is zero. Every real-valued integral appearing here and below is the Bochner integral, which takes the value zero if its integrand is not integrable. For any real-valued function on the state space, let be the measure obtained from by the density ; thus for measurable . The theorem asserts the conjunction of the following two claims. First, for every positive integer , if and is symmetric positive definite, then agrees -almost everywhere with a strongly measurable function, , the pointwise identity holds for every , and is nonnegative -almost everywhere, is integrable with respect to , and satisfies . This first claim is conditional separately at each index and imposes no conclusion at an index where its matrix conditions fail. Second, if for every bounded continuous real-valued function on both
then there is a positive integer such that, for every positive integer , all of the following hold: ; is symmetric positive definite; agrees -almost everywhere with a strongly measurable function; ; is nonnegative -almost everywhere, is -integrable, and has integral one; and there exist a vector and a symmetric positive definite real matrix such that for -almost every . These existential Gaussian parameters need not be stated as unique. In addition, for every bounded continuous real-valued function on ,
Every sequence limit is along the positive integers; there is no index zero. The convergence hypothesis concerning requires only that it be a probability measure, with no density or Gaussian assumption. The second claim gives the stated matrix and normalization properties only eventually; the full sequence in its last limit still uses the total formula for at every positive index, including any earlier index with or , when that formula gives .
Confirmed by the mission captain (proposal self-audit).