Stat.Limit.ObservationDependentVanTrees
Observation-dependent Bayesian Cramér–Rao bounds: guarded-score algebra, absolute-continuity integration by parts, product-measure identities, and weighted L² Cauchy–Schwarz for dominated likelihood models.
Basic 15 core · 16 supporting This module defines the restricted parameter measure, guarded prior, likelihood, and joint scores, and the weighted fields used by the product-measure proof. ★ priorInformation_smoothPrior
Basic fields for observation-dependent van Trees bounds
This module defines the restricted parameter measure, guarded prior, likelihood, and joint scores, and the weighted fields used by the product-measure proof. The guards make every definition meaningful where a density vanishes.
Given a lower endpoint and an upper endpoint, the parameter reference measure is Lebesgue measure restricted to the closed interval from the lower to the upper endpoint.
Definition (Lean source)
Given a prior density on the real parameter, a conditional observation density given that parameter, and a parameter--observation pair, the joint density is the product of the prior density and the conditional likelihood density at that pair.
Definition (Lean source)
Given a prior density, its parameter derivative, and a parameter value, the guarded prior score is the derivative divided by the density when the density is positive, and is zero otherwise.
Definition (Lean source)
Given a conditional likelihood density, its parameter derivative, a parameter value, and an observation, the guarded likelihood score is the derivative divided by the likelihood when that likelihood is positive, and is zero otherwise.
Definition (Lean source)
Given a prior density, its derivative, a conditional likelihood density, its derivative, and a parameter--observation pair, the guarded joint score is the derivative of the prior--likelihood product divided by that product when the product is positive, and is zero otherwise.
Definition (Lean source)
Given a lower endpoint, an upper endpoint, a prior density, and its derivative, the prior Fisher information is the integral over the closed parameter interval of the prior density times the squared guarded prior score.
Definition (Lean source)
Given a measure on a measurable observation space, a conditional likelihood density, its parameter derivative, and a parameter value, the conditional Fisher information is the observation integral of the likelihood times the squared guarded likelihood score at that parameter.
Definition (Lean source)
Given a prior density, its derivative, a conditional likelihood density, its derivative, a target function, an estimator, and a parameter--observation pair, the signed error--joint-score field is estimator error times guarded joint score, weighted by joint density.
Definition (Lean source)
Given a prior density, a conditional likelihood density, the parameter derivative of a target function, and a parameter--observation pair, the target-sensitivity field is that derivative weighted by the joint density.
Definition (Lean source)
Given a prior density, its derivative, a conditional likelihood density, its derivative, a target function, its parameter derivative, an estimator, and a parameter--observation pair, the derivative-balance field is the parameter derivative of the joint-density-weighted estimation error.
Definition (Lean source)
Given a prior density, a conditional likelihood density, a target function, an estimator, and a parameter--observation pair, the squared-error field is the squared estimator error weighted by the joint density.
Definition (Lean source)
Given a prior density, its derivative, a conditional likelihood density, its derivative, and a parameter--observation pair, the squared-score field is the squared guarded joint score weighted by the joint density.
Definition (Lean source)
Given a center, a radius, and a parameter value, the smooth quartic prior is when , and is zero otherwise.
Definition (Lean source)
Given a center, a radius, and a parameter value, the smooth quartic prior derivative is when , and is zero otherwise.
Definition (Lean source)
If the bandwidth is positive and the prior's left and right support endpoints lie in an ambient interval, then the prior Fisher information equals ten divided by the squared bandwidth.
Formal statement
Proof (Lean source)
16 supporting declarations (lemmas, instances)
-
smoothPrior_nonnegtheorem — A positive bandwidth makes the smooth quartic prior nonnegative at every parameter value.Proof (Lean source)
theorem smoothPrior_nonneg {c a : ℝ} (ha : 0 < a) (θ : ℝ) : 0 ≤ smoothPrior c a θ := by unfold smoothPrior split_ifs <;> positivity -
smoothPrior_pos_ifftheorem — Under a positive bandwidth, the smooth quartic prior is positive exactly inside the open interval defined by its center and bandwidth.Proof (Lean source)
theorem smoothPrior_pos_iff {c a θ : ℝ} (ha : 0 < a) : 0 < smoothPrior c a θ ↔ |θ - c| < a := by unfold smoothPrior split_ifs with h · simp only [h, iff_true] have hr : |(θ - c) / a| < 1 := by rw [abs_div, abs_of_pos ha, div_lt_one ha] exact h have hs : ((θ - c) / a) ^ 2 < 1 := by rw [sq_lt_one_iff_abs_lt_one] exact hr positivity · simp [h] -
support_smoothPriortheorem — Under a positive bandwidth, the nonzero set of the smooth quartic prior is exactly its open support interval.Proof (Lean source)
theorem support_smoothPrior {c a : ℝ} (ha : 0 < a) : support (smoothPrior c a) = Ioo (c - a) (c + a) := by ext θ rw [Function.mem_support, mem_Ioo] constructor · intro hne have hpos : 0 < smoothPrior c a θ := lt_of_le_of_ne (smoothPrior_nonneg ha θ) (Ne.symm hne) have habs := (smoothPrior_pos_iff ha).mp hpos rw [abs_lt] at habs constructor <;> linarith · rintro ⟨hleft, hright⟩ apply ne_of_gt rw [smoothPrior_pos_iff ha, abs_lt] constructor <;> linarith -
tsupport_smoothPriortheorem — Under a positive bandwidth, the topological support of the smooth quartic prior is exactly the corresponding closed interval.Proof (Lean source)
theorem tsupport_smoothPrior {c a : ℝ} (ha : 0 < a) : tsupport (smoothPrior c a) = Icc (c - a) (c + a) := by rw [tsupport, support_smoothPrior ha, closure_Ioo] linarith -
support_smoothPrior_subset_Icctheorem — If the bandwidth is positive and the prior's left and right support endpoints lie strictly inside an ambient interval, then every point where the prior is nonzero lies in the ambient closed interval.hypothesesell u c a :ℝha :0 < ahleft :ell < c - ahright :c + a < uconclusionsupport (smoothPrior c a) ⊆ Icc ell uProof (Lean source)
theorem support_smoothPrior_subset_Icc {ell u c a : ℝ} (ha : 0 < a) (hleft : ell < c - a) (hright : c + a < u) : support (smoothPrior c a) ⊆ Icc ell u := by rw [support_smoothPrior ha] rintro θ ⟨hθleft, hθright⟩ constructor <;> linarith -
tsupport_smoothPrior_subset_Iootheorem — If the bandwidth is positive and the prior's left and right support endpoints lie strictly inside an ambient interval, then the prior's topological support lies in the ambient open interval.hypothesesell u c a :ℝha :0 < ahleft :ell < c - ahright :c + a < uconclusionProof (Lean source)
theorem tsupport_smoothPrior_subset_Ioo {ell u c a : ℝ} (ha : 0 < a) (hleft : ell < c - a) (hright : c + a < u) : tsupport (smoothPrior c a) ⊆ Ioo ell u := by rw [tsupport_smoothPrior ha] rintro θ ⟨hθleft, hθright⟩ constructor <;> linarith -
smoothPrior_ambient_endpointstheorem — If the bandwidth is positive and the prior's left and right support endpoints lie strictly inside an ambient interval, then the prior vanishes at both ambient endpoints.hypothesesell u c a :ℝha :0 < ahleft :ell < c - ahright :c + a < uProof (Lean source)
theorem smoothPrior_ambient_endpoints {ell u c a : ℝ} (ha : 0 < a) (hleft : ell < c - a) (hright : c + a < u) : smoothPrior c a ell = 0 ∧ smoothPrior c a u = 0 := by constructor · unfold smoothPrior rw [if_neg] rw [abs_lt] intro h linarith · unfold smoothPrior rw [if_neg] rw [abs_lt] intro h linarith -
smoothPrior_contDifftheorem — A positive bandwidth makes the smooth quartic prior continuously differentiable.Proof (Lean source)
theorem smoothPrior_contDiff {c a : ℝ} (ha : 0 < a) : ContDiff ℝ 1 (smoothPrior c a) := by rw [contDiff_one_iff_deriv] constructor · intro θ exact (hasDerivAt_smoothPrior_aux ha θ).differentiableAt · have hderiv : deriv (smoothPrior c a) = smoothPriorDeriv c a := by funext θ exact (hasDerivAt_smoothPrior_aux ha θ).deriv rw [hderiv] exact continuous_smoothPriorDeriv ha -
hasDerivAt_smoothPriortheorem — Under a positive bandwidth, the explicit derivative representative is the derivative of the smooth quartic prior at every parameter value.hypothesesc a :ℝha :0 < aθ :ℝconclusionProof (Lean source)
theorem hasDerivAt_smoothPrior {c a : ℝ} (ha : 0 < a) (θ : ℝ) : HasDerivAt (smoothPrior c a) (smoothPriorDeriv c a θ) θ := by exact hasDerivAt_smoothPrior_aux ha θ -
smoothPrior_absolutelyContinuousOnIntervaltheorem — A positive bandwidth and ordered endpoints ensure that the smooth quartic prior is absolutely continuous on the interval.hypothesesell u c a :ℝha :0 < ahellu :ell ≤ uconclusionAbsolutelyContinuousOnInterval (smoothPrior c a) ell uProof (Lean source)
theorem smoothPrior_absolutelyContinuousOnInterval {ell u c a : ℝ} (ha : 0 < a) (hellu : ell ≤ u) : AbsolutelyContinuousOnInterval (smoothPrior c a) ell u := by exact (smoothPrior_contDiff ha).contDiffOn.absolutelyContinuousOnInterval -
integral_smoothPrior_volumetheorem — Under a positive bandwidth, the smooth quartic density has total Lebesgue integral one.Proof (Lean source)
theorem integral_smoothPrior_volume {c a : ℝ} (ha : 0 < a) : ∫ θ, smoothPrior c a θ = 1 := by let p : ℝ → ℝ := fun t => (15 / (16 * a)) * (1 - ((t - c) / a) ^ 2) ^ 2 have hinterval : (∫ t in c - a..c + a, p t) = 1 := by let f : ℝ → ℝ := fun x => (1 - x ^ 2) ^ 2 have hf : (∫ x : ℝ in (-1)..1, f x) = 16 / 15 := by have hfun : f = fun x => (1 - 2 * x ^ 2) + x ^ 4 := by funext x dsimp [f] ring rw [hfun, intervalIntegral.integral_add (Continuous.intervalIntegrable (μ := volume) (by fun_prop : Continuous (fun x : ℝ => 1 - 2 * x ^ 2)) (-1) 1) (Continuous.intervalIntegrable (μ := volume) (by fun_prop : Continuous (fun x : ℝ => x ^ 4)) (-1) 1)] rw [intervalIntegral.integral_sub (Continuous.intervalIntegrable (μ := volume) (by fun_prop : Continuous (fun _ : ℝ => (1 : ℝ))) (-1) 1) (Continuous.intervalIntegrable (μ := volume) (by fun_prop : Continuous (fun x : ℝ => 2 * x ^ 2)) (-1) 1)] rw [intervalIntegral.integral_const_mul, integral_pow, integral_pow] norm_num dsimp [p] rw [intervalIntegral.integral_const_mul] have hcomp := intervalIntegral.integral_comp_div_sub (a := c - a) (b := c + a) f ha.ne' (c / a) have heq : (∫ t in c - a..c + a, f ((t - c) / a)) = a * (∫ x in (-1)..1, f x) := by rw [show (fun t : ℝ => f ((t - c) / a)) = fun t => f (t / a - c / a) by funext t congr 1 field_simp [ha.ne']] rw [hcomp] congr 2 <;> field_simp [ha.ne'] <;> ring rw [show (fun t : ℝ => (1 - ((t - c) / a) ^ 2) ^ 2) = fun t => f ((t - c) / a) by rfl] rw [heq, hf] field_simp [ha.ne'] have hfun : smoothPrior c a = (Ioo (c - a) (c + a)).indicator p := by funext t by_cases ht : |t - c| < a · have hmem : t ∈ Ioo (c - a) (c + a) := by rw [mem_Ioo] rw [abs_lt] at ht constructor <;> linarith simp [smoothPrior, p, ht, hmem] · have hnotmem : t ∉ Ioo (c - a) (c + a) := by rw [mem_Ioo] rintro ⟨hlt, hrt⟩ apply ht rw [abs_lt] constructor <;> linarith simp [smoothPrior, ht, hnotmem] rw [hfun, integral_indicator measurableSet_Ioo] rw [← integral_Ioc_eq_integral_Ioo] rw [← intervalIntegral.integral_of_le (by linarith : c - a ≤ c + a)] exact hinterval -
integral_smoothPrior_parameterMeasuretheorem — If the bandwidth is positive and the prior's left and right support endpoints lie in an ambient interval, then the prior has mass one under Lebesgue measure restricted to that interval.hypothesesell u c a :ℝha :0 < ahleft :ell ≤ c - ahright :c + a ≤ uconclusion∫ θ, smoothPrior c a θ ∂parameterMeasure ell u = 1Proof (Lean source)
theorem integral_smoothPrior_parameterMeasure {ell u c a : ℝ} (ha : 0 < a) (hleft : ell ≤ c - a) (hright : c + a ≤ u) : ∫ θ, smoothPrior c a θ ∂parameterMeasure ell u = 1 := by have hsupport : support (smoothPrior c a) ⊆ Icc ell u := by rw [support_smoothPrior ha] rintro θ ⟨hθl, hθr⟩ exact ⟨hleft.trans hθl.le, hθr.le.trans hright⟩ unfold parameterMeasure rw [← integral_indicator measurableSet_Icc] have hfun : (Icc ell u).indicator (smoothPrior c a) = smoothPrior c a := by funext θ by_cases hθ : θ ∈ Icc ell u · simp [hθ] · have hz : smoothPrior c a θ = 0 := by by_contra hnz exact hθ (hsupport hnz) simp [hθ, hz] rw [hfun, integral_smoothPrior_volume ha] -
smoothPrior_integrable_parameterMeasuretheorem — A positive bandwidth makes the smooth quartic prior integrable on every restricted parameter interval.hypothesesell u c a :ℝha :0 < aconclusionProof (Lean source)
theorem smoothPrior_integrable_parameterMeasure {ell u c a : ℝ} (ha : 0 < a) : Integrable (smoothPrior c a) (parameterMeasure ell u) := by unfold parameterMeasure exact (smoothPrior_contDiff ha).continuous.integrableOn_Icc -
smoothPrior_scoreSq_aestronglyMeasurabletheorem — A positive bandwidth makes the prior-weighted squared guarded score strongly measurable on every restricted parameter interval.hypothesesell u c a :ℝha :0 < aconclusionAEStronglyMeasurable (fun θ => smoothPrior c a θ * (priorScore (smoothPrior c a) (smoothPriorDeriv c a) θ) ^ 2) (parameterMeasure ell u)Proof (Lean source)
theorem smoothPrior_scoreSq_aestronglyMeasurable {ell u c a : ℝ} (ha : 0 < a) : AEStronglyMeasurable (fun θ => smoothPrior c a θ * (priorScore (smoothPrior c a) (smoothPriorDeriv c a) θ) ^ 2) (parameterMeasure ell u) := by exact (smoothPrior_scoreSq_integrable_aux ha).aestronglyMeasurable -
smoothPrior_scoreSq_integrabletheorem — A positive bandwidth makes the prior-weighted squared guarded score integrable on every restricted parameter interval.hypothesesell u c a :ℝha :0 < aconclusionIntegrable (fun θ => smoothPrior c a θ * (priorScore (smoothPrior c a) (smoothPriorDeriv c a) θ) ^ 2) (parameterMeasure ell u)Proof (Lean source)
theorem smoothPrior_scoreSq_integrable {ell u c a : ℝ} (ha : 0 < a) : Integrable (fun θ => smoothPrior c a θ * (priorScore (smoothPrior c a) (smoothPriorDeriv c a) θ) ^ 2) (parameterMeasure ell u) := by exact smoothPrior_scoreSq_integrable_aux ha -
priorInformation_smoothPrior_letheorem — If the bandwidth is positive and the prior's left and right support endpoints lie in an ambient interval, then the prior Fisher information is at most forty divided by the squared bandwidth.hypothesesell u c a :ℝha :0 < ahleft :ell ≤ c - ahright :c + a ≤ uconclusionProof (Lean source)
theorem priorInformation_smoothPrior_le {ell u c a : ℝ} (ha : 0 < a) (hleft : ell ≤ c - a) (hright : c + a ≤ u) : priorInformation ell u (smoothPrior c a) (smoothPriorDeriv c a) ≤ 40 / a ^ 2 := by rw [priorInformation_smoothPrior ha hleft hright] exact (div_le_div_iff_of_pos_right (sq_pos_of_pos ha)).2 (by norm_num)
GuardedInformation 2 core · 12 supporting This module proves the density-zero algebra for guarded scores, derives score centering from likelihood normalization and differentiation under the integral, and decomposes the joint score information into prior and Fish ★ joint_score_information_decomposition★ average_fisherInformation_le
Guarded scores, centering, and information decomposition
This module proves the density-zero algebra for guarded scores, derives score centering from likelihood normalization and differentiation under the integral, and decomposes the joint score information into prior and Fisher terms.
For nonnegative prior and likelihood densities with zero-set derivative control, likelihood normalization, centered conditional scores, and the stated measurable integrability of all square and cross fields, joint-score information equals prior information plus average conditional Fisher information.
Formal statement
Proof (Lean source)
If the prior is nonnegative, normalized, and integrable, prior-weighted Fisher information is integrable, and conditional Fisher information is pointwise bounded wherever the prior is positive, then average Fisher information is bounded by the same constant.
Formal statement
Proof (Lean source)
12 supporting declarations (lemmas, instances)
-
derivative_eq_zero_of_nonnegative_of_eq_zerotheorem — A differentiable nonnegative function has zero derivative at any point where it attains value zero, so the supplied derivative is zero.hypothesesconclusiondf = 0Proof (Lean source)
theorem derivative_eq_zero_of_nonnegative_of_eq_zero {f : ℝ → ℝ} {df θ : ℝ} (hnonneg : ∀ t, 0 ≤ f t) (hderiv : HasDerivAt f df θ) (hzero : f θ = 0) : df = 0 := by have hmin : IsLocalMin f θ := by filter_upwards with t rw [hzero] exact hnonneg t exact IsLocalMin.hasDerivAt_eq_zero hmin hderiv -
guarded_score_mul_densitytheorem — A nonnegative density with a derivative numerator that vanishes whenever the density vanishes has its guarded score times its density equal to that derivative numerator.hypothesesq dq :ℝhq :0 ≤ qhzero :q = 0 → dq = 0conclusion(if 0 < q then dq / q else 0) * q = dqProof (Lean source)
theorem guarded_score_mul_density {q dq : ℝ} (hq : 0 ≤ q) (hzero : q = 0 → dq = 0) : (if 0 < q then dq / q else 0) * q = dq := by by_cases hqpos : 0 < q · simp [hqpos, ne_of_gt hqpos] · have hqzero : q = 0 := le_antisymm (le_of_not_gt hqpos) hq simp [hqzero, hzero hqzero] -
jointScore_eq_addtheorem — At a point where both the prior and likelihood densities are positive, the guarded joint score equals the sum of their guarded scores.hypothesesX :Type*w dw :ℝ → ℝp dp :ℝ → X → ℝθ :ℝx :Xhw :0 < w θhp :0 < p θ xconclusionProof (Lean source)
theorem jointScore_eq_add {X : Type*} {w dw : ℝ → ℝ} {p dp : ℝ → X → ℝ} {θ : ℝ} {x : X} (hw : 0 < w θ) (hp : 0 < p θ x) : jointScore w dw p dp (θ, x) = priorScore w dw θ + likelihoodScore p dp θ x := by simp only [jointScore, jointDensity, priorScore, likelihoodScore, hw, hp, mul_pos hw hp, if_true] field_simp -
jointScore_mul_jointDensitytheorem — For nonnegative prior and likelihood densities whose derivative numerators vanish on their respective zero-density sets, joint density times guarded joint score equals the product-rule derivative numerator.hypothesesX :Type*w dw :ℝ → ℝp dp :ℝ → X → ℝθ :ℝx :Xhw :0 ≤ w θhp :0 ≤ p θ xhwzero :w θ = 0 → dw θ = 0hpzero :p θ x = 0 → dp θ x = 0conclusionjointScore w dw p dp (θ, x) * jointDensity w p (θ, x) = dw θ * p θ x + w θ * dp θ xProof (Lean source)
theorem jointScore_mul_jointDensity {X : Type*} {w dw : ℝ → ℝ} {p dp : ℝ → X → ℝ} {θ : ℝ} {x : X} (hw : 0 ≤ w θ) (hp : 0 ≤ p θ x) (hwzero : w θ = 0 → dw θ = 0) (hpzero : p θ x = 0 → dp θ x = 0) : jointScore w dw p dp (θ, x) * jointDensity w p (θ, x) = dw θ * p θ x + w θ * dp θ x := by apply guarded_score_mul_density (mul_nonneg hw hp) intro hzero rcases mul_eq_zero.mp hzero with hwz | hpz · rw [hwzero hwz, hwz] ring · rw [hpzero hpz, hpz] ring -
errorScoreField_eq_numeratortheorem — For nonnegative prior and likelihood densities whose derivative numerators vanish wherever the corresponding density is zero, the weighted error--score field equals estimation error times the unguarded product-rule numerator.hypothesesX :Type*w dw :ℝ → ℝp dp g :ℝ → X → ℝT :X → ℝθ :ℝx :Xhw :0 ≤ w θhp :0 ≤ p θ xhwzero :w θ = 0 → dw θ = 0hpzero :p θ x = 0 → dp θ x = 0conclusionerrorScoreField w dw p dp g T (θ, x) = (T x - g θ x) * (dw θ * p θ x + w θ * dp θ x)Proof (Lean source)
theorem errorScoreField_eq_numerator {X : Type*} {w dw : ℝ → ℝ} {p dp g : ℝ → X → ℝ} {T : X → ℝ} {θ : ℝ} {x : X} (hw : 0 ≤ w θ) (hp : 0 ≤ p θ x) (hwzero : w θ = 0 → dw θ = 0) (hpzero : p θ x = 0 → dp θ x = 0) : errorScoreField w dw p dp g T (θ, x) = (T x - g θ x) * (dw θ * p θ x + w θ * dp θ x) := by have hjoint := jointScore_mul_jointDensity (w := w) (dw := dw) (p := p) (dp := dp) (θ := θ) (x := x) hw hp hwzero hpzero rw [errorScoreField, mul_assoc, hjoint] -
likelihoodScore_integral_eq_zero_of_normalizationtheorem — At an interior parameter value, a normalized likelihood whose integral may be differentiated using the supplied derivative, with nonnegative density, zero-set derivative control, and both fields integrable has conditional mean-zero guarded likelihood score.hypothesesX :a b θ :ℝp dp :ℝ → X → ℝhθ :θ ∈ Ioo a bhnorm :∀ t ∈ Icc a b, ∫ x, p t x ∂μ = 1hdiff :HasDerivAt (fun t => ∫ x, p t x ∂μ) (∫ x, dp θ x ∂μ) θhp :∀ x, 0 ≤ p θ xhpzero :∀ᵐ x ∂μ, p θ x = 0 → dp θ x = 0hpint :Integrable (fun x => p θ x) μhdpint :Integrable (fun x => dp θ x) μconclusion∫ x, likelihoodScore p dp θ x * p θ x ∂μ = 0Proof (Lean source)
theorem likelihoodScore_integral_eq_zero_of_normalization {X : Type*} [MeasurableSpace X] {μ : Measure X} {a b θ : ℝ} {p dp : ℝ → X → ℝ} (hθ : θ ∈ Ioo a b) (hnorm : ∀ t ∈ Icc a b, ∫ x, p t x ∂μ = 1) (hdiff : HasDerivAt (fun t => ∫ x, p t x ∂μ) (∫ x, dp θ x ∂μ) θ) (hp : ∀ x, 0 ≤ p θ x) (hpzero : ∀ᵐ x ∂μ, p θ x = 0 → dp θ x = 0) (hpint : Integrable (fun x => p θ x) μ) (hdpint : Integrable (fun x => dp θ x) μ) : ∫ x, likelihoodScore p dp θ x * p θ x ∂μ = 0 := by have hnorm_nhds : (fun t => ∫ x, p t x ∂μ) =ᶠ[nhds θ] fun _ => 1 := by filter_upwards [Ioo_mem_nhds hθ.1 hθ.2] with t ht exact hnorm t ⟨ht.1.le, ht.2.le⟩ have hderiv_zero : (∫ x, dp θ x ∂μ) = 0 := by apply hdiff.unique exact (hasDerivAt_const θ 1).congr_of_eventuallyEq hnorm_nhds have hscore_mul : (fun x => likelihoodScore p dp θ x * p θ x) =ᵐ[μ] fun x => dp θ x := by filter_upwards [hpzero] with x hx exact guarded_score_mul_density (hp x) hx rw [integral_congr_ae hscore_mul, hderiv_zero] -
finite_likelihood_normalizationtheorem — If the masses of a finite likelihood sum to one, then its integral against counting measure is one.hypothesesconclusion∫ x, p θ x ∂Measure.count = 1Proof (Lean source)
theorem finite_likelihood_normalization {X : Type*} [Fintype X] [MeasurableSpace X] [MeasurableSingletonClass X] {p : ℝ → X → ℝ} {θ : ℝ} (hnorm : ∑ x, p θ x = 1) : ∫ x, p θ x ∂Measure.count = 1 := by rw [integral_count, hnorm] -
hasDerivAt_finite_likelihood_sumtheorem — If each cell probability has the supplied parameter derivative, then the derivative of their finite sum is the sum of those derivatives.hypothesesconclusionHasDerivAt (fun t => ∑ x, p t x) (∑ x, dp θ x) θProof (Lean source)
theorem hasDerivAt_finite_likelihood_sum {X : Type*} [Fintype X] {p dp : ℝ → X → ℝ} {θ : ℝ} (hderiv : ∀ x, HasDerivAt (fun t => p t x) (dp θ x) θ) : HasDerivAt (fun t => ∑ x, p t x) (∑ x, dp θ x) θ := by have hsum : HasDerivAt (∑ x : X, fun t => p t x) (∑ x, dp θ x) θ := HasDerivAt.sum (u := univ) (fun x _ => hderiv x) have heq : (fun t => ∑ x, p t x) = (∑ x : X, fun t => p t x) := by funext t rw [Finset.sum_apply] rw [heq] exact hsum -
finite_derivative_centeringtheorem — If each cell probability has the supplied derivative and the finite likelihood is locally normalized, then the derivative masses sum to zero.hypothesesX :θ :ℝhderiv :∀ x, HasDerivAt (fun t => p t x) (dp θ x) θhnorm :∀ᶠ t in nhds θ, ∑ x, p t x = 1conclusion∑ x, dp θ x = 0Proof (Lean source)
theorem finite_derivative_centering {X : Type*} [Fintype X] {p dp : ℝ → X → ℝ} {θ : ℝ} (hderiv : ∀ x, HasDerivAt (fun t => p t x) (dp θ x) θ) (hnorm : ∀ᶠ t in nhds θ, ∑ x, p t x = 1) : ∑ x, dp θ x = 0 := by have hsum := hasDerivAt_finite_likelihood_sum hderiv apply hsum.unique exact (hasDerivAt_const θ 1).congr_of_eventuallyEq hnorm -
finite_likelihoodScore_multheorem — If finite likelihood masses are nonnegative and their derivatives vanish on zero-mass cells, then likelihood mass times guarded score equals the cell derivative.hypothesesX :Type*ℝ → X → ℝθ :ℝhp :∀ x, 0 ≤ p θ xhzero :∀ xifp θ x = 0thendp θ x = 0x :XconclusionlikelihoodScore p dp θ x * p θ x = dp θ xProof (Lean source)
theorem finite_likelihoodScore_mul {X : Type*} {p dp : ℝ → X → ℝ} {θ : ℝ} (hp : ∀ x, 0 ≤ p θ x) (hzero : ∀ x, p θ x = 0 → dp θ x = 0) (x : X) : likelihoodScore p dp θ x * p θ x = dp θ x := by exact guarded_score_mul_density (hp x) (hzero x) -
finite_likelihoodScore_centeredtheorem — If finite likelihood masses are nonnegative, derivatives vanish on zero-mass cells, and the derivative masses are centered, then the likelihood-weighted guarded scores sum to zero.hypothesesX :θ :ℝhp :∀ x, 0 ≤ p θ xhzero :∀ xifp θ x = 0thendp θ x = 0hcenter :∑ x, dp θ x = 0conclusion∑ x, likelihoodScore p dp θ x * p θ x = 0Proof (Lean source)
theorem finite_likelihoodScore_centered {X : Type*} [Fintype X] {p dp : ℝ → X → ℝ} {θ : ℝ} (hp : ∀ x, 0 ≤ p θ x) (hzero : ∀ x, p θ x = 0 → dp θ x = 0) (hcenter : ∑ x, dp θ x = 0) : ∑ x, likelihoodScore p dp θ x * p θ x = 0 := by simpa only [finite_likelihoodScore_mul hp hzero] using hcenter -
fisherInformation_count_eq_sumtheorem — Conditional Fisher information under counting measure equals the finite sum of likelihood mass times squared guarded score.hypothesesconclusionfisherInformation Measure.count p dp θ = ∑ x, p θ x * (likelihoodScore p dp θ x) ^ 2Proof (Lean source)
theorem fisherInformation_count_eq_sum {X : Type*} [Fintype X] [MeasurableSpace X] [MeasurableSingletonClass X] (p dp : ℝ → X → ℝ) (θ : ℝ) : fisherInformation Measure.count p dp θ = ∑ x, p θ x * (likelihoodScore p dp θ x) ^ 2 := by exact integral_count _
IntegrationByParts 1 core · 2 supporting This module supplies integration-by-parts identities from absolute continuity and almost-everywhere derivative representatives, including the weighted-error identity used for observation-dependent targets. ★ product_integral_derivativeBalance_eq_zero
Absolute-continuity integration by parts
This module supplies integration-by-parts identities from absolute continuity and almost-everywhere derivative representatives, including the weighted-error identity used for observation-dependent targets.
In a σ-finite observation model, if the prior weight and almost every likelihood and target section are absolutely continuous with the stated derivative representatives, endpoint error products vanish, and the derivative-balance field is measurable and integrable, the joint integral of the derivative-balance field is zero.
Formal statement
Proof (Lean source)
2 supporting declarations (lemmas, instances)
-
ac_product_integral_eq_boundarytheorem — Two functions on an ordered interval with absolute continuity, almost-everywhere derivative representatives, and an integrable product-rule field satisfy the product-rule integral equals the endpoint change in their product.hypothesesa b :ℝf g df dg :ℝ → ℝhab :a ≤ bhf :hg :hdf :∀ᵐ θ ∂parameterMeasure a b, HasDerivAt f (df θ) θhdg :∀ᵐ θ ∂parameterMeasure a b, HasDerivAt g (dg θ) θhint :IntervalIntegrable (fun θ => df θ * g θ + f θ * dg θ) volume a bconclusion(∫ θ in a..b, df θ * g θ + f θ * dg θ) = f b * g b - f a * g aProof (Lean source)
theorem ac_product_integral_eq_boundary {a b : ℝ} {f g df dg : ℝ → ℝ} (hab : a ≤ b) (hf : AbsolutelyContinuousOnInterval f a b) (hg : AbsolutelyContinuousOnInterval g a b) (hdf : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt f (df θ) θ) (hdg : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt g (dg θ) θ) (hint : IntervalIntegrable (fun θ => df θ * g θ + f θ * dg θ) volume a b) : (∫ θ in a..b, df θ * g θ + f θ * dg θ) = f b * g b - f a * g a := by let q : ℝ → ℝ := fun θ => f θ * g θ have hq : AbsolutelyContinuousOnInterval q a b := hf.fun_mul hg have hqderiv : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt q (df θ * g θ + f θ * dg θ) θ := by filter_upwards [hdf, hdg] with θ hdfθ hdgθ exact hdfθ.mul hdgθ calc ∫ θ in a..b, df θ * g θ + f θ * dg θ = ∫ θ, df θ * g θ + f θ * dg θ ∂parameterMeasure a b := by rw [intervalIntegral.integral_of_le hab, parameterMeasure, restrict_Ioc_eq_restrict_Icc] _ = ∫ θ, deriv q θ ∂parameterMeasure a b := by apply MeasureTheory.integral_congr_ae filter_upwards [hqderiv] with θ hθ exact hθ.deriv.symm _ = ∫ θ in a..b, deriv q θ := by rw [intervalIntegral.integral_of_le hab, parameterMeasure, restrict_Ioc_eq_restrict_Icc] _ = q b - q a := hq.integral_deriv_eq_sub _ = f b * g b - f a * g a := rfl -
ac_weighted_error_integral_eq_boundarytheorem — A prior weight, likelihood section, and observation-dependent target with absolute continuity, almost-everywhere derivative representatives, and an integrable weighted-error derivative field satisfy the weighted-error derivative integral equals the weighted endpoint difference.hypothesesa b c :ℝw p g dw dp dg :ℝ → ℝhab :a ≤ bhw :hp :hg :hdw :∀ᵐ θ ∂parameterMeasure a b, HasDerivAt w (dw θ) θhdp :∀ᵐ θ ∂parameterMeasure a b, HasDerivAt p (dp θ) θhdg :∀ᵐ θ ∂parameterMeasure a b, HasDerivAt g (dg θ) θhint :IntervalIntegrable (fun θ => (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ) volume a bconclusion(∫ θ in a..b, (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ)= w b * p b * (c - g b) - w a * p a * (c - g a)Proof (Lean source)
theorem ac_weighted_error_integral_eq_boundary {a b c : ℝ} {w p g dw dp dg : ℝ → ℝ} (hab : a ≤ b) (hw : AbsolutelyContinuousOnInterval w a b) (hp : AbsolutelyContinuousOnInterval p a b) (hg : AbsolutelyContinuousOnInterval g a b) (hdw : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt w (dw θ) θ) (hdp : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt p (dp θ) θ) (hdg : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt g (dg θ) θ) (hint : IntervalIntegrable (fun θ => (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ) volume a b) : (∫ θ in a..b, (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ) = w b * p b * (c - g b) - w a * p a * (c - g a) := by let q : ℝ → ℝ := fun θ => w θ * p θ * (c - g θ) have hconst : AbsolutelyContinuousOnInterval (fun _ : ℝ => c) a b := by rw [absolutelyContinuousOnInterval_iff] intro ε hε exact ⟨1, zero_lt_one, by intros; simpa using hε⟩ have hq : AbsolutelyContinuousOnInterval q a b := (hw.fun_mul hp).fun_mul (hconst.sub hg) have hqderiv : ∀ᵐ θ ∂parameterMeasure a b, HasDerivAt q ((dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ) θ := by filter_upwards [hdw, hdp, hdg] with θ hdwθ hdpθ hdgθ have hmul := (hdwθ.mul hdpθ).mul (hdgθ.const_sub c) have hder : (dw θ * p θ + w θ * dp θ) * (c - g θ) + (w * p) θ * -dg θ = (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ := by simp only [Pi.mul_apply] ring rw [hder] at hmul exact hmul calc ∫ θ in a..b, (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ = ∫ θ, (dw θ * p θ + w θ * dp θ) * (c - g θ) - w θ * p θ * dg θ ∂parameterMeasure a b := by rw [intervalIntegral.integral_of_le hab, parameterMeasure, restrict_Ioc_eq_restrict_Icc] _ = ∫ θ, deriv q θ ∂parameterMeasure a b := by apply MeasureTheory.integral_congr_ae filter_upwards [hqderiv] with θ hθ exact hθ.deriv.symm _ = ∫ θ in a..b, deriv q θ := by rw [intervalIntegral.integral_of_le hab, parameterMeasure, restrict_Ioc_eq_restrict_Icc] _ = q b - q a := hq.integral_deriv_eq_sub _ = w b * p b * (c - g b) - w a * p a * (c - g a) := rfl
Main 3 core · 1 supporting This module proves a paper-independent Bayesian Cramér--Rao inequality for a sigma-finite dominated observation model. ★ observation_dependent_van_trees★ smoothPrior_finite_vanTrees_lower_bound
Observation-dependent van Trees inequality
This module proves a paper-independent Bayesian Cramér--Rao inequality for a sigma-finite dominated observation model. Unlike the classical parameter-only form, the target may depend on both the parameter and the observation. It assembles the guarded-score, absolute-continuity, product-measure, and weighted L² ingredients from the accompanying modules.
A nondegenerate parameter interval with a continuously differentiable, compactly supported, normalized nonnegative prior; a normalized nonnegative dominated likelihood; the stated differentiation-under-the-integral, sectionwise absolute-continuity, derivative, boundary, product-measurability, and integrability conditions; and strictly positive finite total information ensures that Bayes mean-squared error for an estimator of an observation-dependent target is at least the squared joint mean of the target's parameter derivative divided by prior information plus average Fisher information.
Formal statement
Proof (Lean source)
Finite-experiment van Trees regularity collects the model-specific normalization, absolute-continuity, derivative, measurability, and integrability conditions for a fixed prior and counting-measure observation model.
Definition (Lean source)
Every likelihood mass is nonnegative.
Every likelihood mass function is normalized on the ambient interval.
Likelihood sections are integrable on the finite carrier.
Likelihood derivative sections are integrable on the finite carrier.
Differentiation of the normalized likelihood integral is valid in the interior.
Almost every likelihood section is absolutely continuous in the parameter.
Almost every observation-dependent target section is absolutely continuous.
The supplied likelihood derivative is valid almost everywhere jointly.
The supplied target derivative is valid almost everywhere jointly.
The derivative-balance field is strongly measurable.
The derivative-balance field is integrable.
The error-score field is strongly measurable.
The error-score field is integrable.
The sensitivity field is strongly measurable.
The sensitivity field is integrable.
The weighted squared-error field is strongly measurable.
The weighted squared-error field is integrable.
The joint squared-score field is strongly measurable.
The joint squared-score field is integrable.
The lifted prior-score square is strongly measurable.
The lifted prior-score square is integrable.
The lifted likelihood-score square is strongly measurable.
The lifted likelihood-score square is integrable.
The prior-likelihood score cross field is strongly measurable.
The prior-likelihood score cross field is integrable.
A positive bandwidth, strict containment of the left and right prior-support endpoints, finite-model regularity, a nonnegative sensitivity bound that lies below average sensitivity, a pointwise likelihood-information bound, and positive total information imply a native-real Bayes squared-risk lower bound equal to squared sensitivity divided by likelihood information plus forty over the squared bandwidth.
Formal statement
Proof (Lean source)
1 supporting declaration (lemmas, instances)
-
finite_vanTrees_lower_boundtheorem — A nondegenerate parameter interval, [finite-model regularity] (hyp:M), continuous differentiability, compact support, a valid prior derivative, nonnegativity, normalization, vanishing boundary errors, measurability and integrability of prior information, a nonnegative sensitivity bound that lies below average sensitivity, upper bounds on average likelihood information and prior information, and positive total information imply a native-real Bayes squared-risk lower bound equal to squared sensitivity divided by the sum of the two information bounds.hypothesesX :ell u s I P :ℝw dw :ℝ → ℝp dp g dg :ℝ → X → ℝT :X → ℝhellu :ell < uM :FiniteVanTreesModelRegularity X ell u w dw p dp g dg ThwC1 :ContDiff ℝ 1 whwsupport :support w ⊆ Icc ell uhwderiv :∀ θ, HasDerivAt w (dw θ) θhwnonneg :∀ θ, 0 ≤ w θhwnorm :∫ θ, w θ ∂parameterMeasure ell u = 1hboundary :∀ᵐ x ∂Measure.count, w u * p u x * (T x - g u x)= 0 ∧ w ell * p ell x * (T x - g ell x) = 0hpriorSqSm :hpriorSqInt :hs_nonneg :0 ≤ shsensitivity :hlike :(∫ θ, w θ * fisherInformation Measure.count p dp θ ∂parameterMeasure ell u) ≤ Ihprior :priorInformation ell u w dw ≤ PhinfoPos :0 < priorInformation ell u w dw+ ∫ θ, w θ * fisherInformation Measure.count p dp θ ∂parameterMeasure ell uconclusions ^ 2 / (I + P)Proof (Lean source)
theorem finite_vanTrees_lower_bound {X : Type*} [Fintype X] [MeasurableSpace X] [MeasurableSingletonClass X] {ell u s I P : ℝ} {w dw : ℝ → ℝ} {p dp g dg : ℝ → X → ℝ} {T : X → ℝ} (hellu : ell < u) (M : FiniteVanTreesModelRegularity X ell u w dw p dp g dg T) (hwC1 : ContDiff ℝ 1 w) (hwsupport : support w ⊆ Icc ell u) (hwderiv : ∀ θ, HasDerivAt w (dw θ) θ) (hwnonneg : ∀ θ, 0 ≤ w θ) (hwnorm : ∫ θ, w θ ∂parameterMeasure ell u = 1) (hboundary : ∀ᵐ x ∂Measure.count, w u * p u x * (T x - g u x) = 0 ∧ w ell * p ell x * (T x - g ell x) = 0) (hpriorSqSm : AEStronglyMeasurable (fun θ => w θ * (priorScore w dw θ) ^ 2) (parameterMeasure ell u)) (hpriorSqInt : Integrable (fun θ => w θ * (priorScore w dw θ) ^ 2) (parameterMeasure ell u)) (hs_nonneg : 0 ≤ s) (hsensitivity : s ≤ ∫ z, sensitivityField w p dg z ∂((parameterMeasure ell u).prod Measure.count)) (hlike : (∫ θ, w θ * fisherInformation Measure.count p dp θ ∂parameterMeasure ell u) ≤ I) (hprior : priorInformation ell u w dw ≤ P) (hinfoPos : 0 < priorInformation ell u w dw + ∫ θ, w θ * fisherInformation Measure.count p dp θ ∂parameterMeasure ell u) : s ^ 2 / (I + P) ≤ ∫ z, errorSqField w p g T z ∂((parameterMeasure ell u).prod Measure.count) := by have hvanTrees := observation_dependent_van_trees Measure.count hellu w dw p dp g dg T hwC1 hwsupport hwderiv hwnonneg hwnorm M.hpnonneg M.hpnorm M.hpint M.hdpint M.hdiffUnder M.hpAC M.hgAC M.hdp M.hdg hboundary M.hbalanceSm M.hbalanceInt M.herrorScoreSm M.herrorScoreInt M.hsensitivitySm M.hsensitivityInt M.herrorSqSm M.herrorSqInt M.hscoreSqSm M.hscoreSqInt hpriorSqSm hpriorSqInt M.hpriorJointSqSm M.hpriorJointSqInt M.hfisherSqSm M.hfisherSqInt M.hcrossSm M.hcrossInt hinfoPos have hs_sq : s ^ 2 ≤ (∫ z, sensitivityField w p dg z ∂((parameterMeasure ell u).prod Measure.count)) ^ 2 := by nlinarith have hdenom : priorInformation ell u w dw + ∫ θ, w θ * fisherInformation Measure.count p dp θ ∂parameterMeasure ell u ≤ I + P := by linarith exact (div_le_div₀ (sq_nonneg _) hs_sq hinfoPos hdenom).trans hvanTrees
WeightedL2 1 core · 1 supporting This module packages Fubini for explicitly integrable product fields and proves the squared weighted Cauchy--Schwarz inequality used in van Trees arguments. ★ weighted_integral_mul_sq_le
Product-measure weighted L2 Cauchy--Schwarz
This module packages Fubini for explicitly integrable product fields and proves the squared weighted Cauchy--Schwarz inequality used in van Trees arguments.
Under a nonnegative weight and measurable, integrable weighted square and cross-product fields, the squared weighted pairing is bounded by the product of the two weighted second moments.
Formal statement
Proof (Lean source)
1 supporting declaration (lemmas, instances)
-
product_integral_eq_iteratedtheorem — A real-valued product field that is almost-everywhere strongly measurable and integrable under two σ-finite measures has both iterated integrals equal to its product-measure integral.hypothesesconclusionconclusion 1(∫ z, F z ∂(μ.prod ν)) = ∫ a, (∫ b, F (a, b) ∂ν) ∂μconclusion 2(∫ z, F z ∂(μ.prod ν)) = ∫ b, (∫ a, F (a, b) ∂μ) ∂νProof (Lean source)
theorem product_integral_eq_iterated {A B : Type*} [MeasurableSpace A] [MeasurableSpace B] {μ : Measure A} {ν : Measure B} [SigmaFinite μ] [SigmaFinite ν] {F : A × B → ℝ} (hsm : AEStronglyMeasurable F (μ.prod ν)) (hint : Integrable F (μ.prod ν)) : (∫ z, F z ∂(μ.prod ν)) = ∫ a, (∫ b, F (a, b) ∂ν) ∂μ ∧ (∫ z, F z ∂(μ.prod ν)) = ∫ b, (∫ a, F (a, b) ∂μ) ∂ν := by have hint' : Integrable F (μ.prod ν) := ⟨hsm, hint.hasFiniteIntegral⟩ exact ⟨integral_prod F hint', integral_prod_symm F hint'⟩