Documentation

TauCeti.MeasureTheory.Measure.Lebesgue.BrunnMinkowski.Homothety

Equality in Brunn–Minkowski for convex bodies: homothety #

Let K and L be compact convex sets of positive volume in a finite-dimensional Euclidean space and let 0 < θ < 1. The multiplicative Brunn–Minkowski inequality volume K ^ (1 - θ) * volume L ^ θ ≤ volume ((1 - θ) • K + θ • L) is an equality exactly when L is a translate of K. For c, d > 0 and n = finrank ℝ E ≥ 1, the additive inequality c * volume A ^ (1 / n) + d * volume B ^ (1 / n) ≤ volume (c • A + d • B) ^ (1 / n) is an equality exactly when B = v +ᵥ r • A for a vector v and a scalar r > 0. Taking c = 1 - θ and d = θ gives the weighted form, and c = d = 1 the unweighted form.

The proof completes the argument of Kneser and Süss. Under multiplicative equality, the height quantiles of K and L along each unit vector u differ by a constant (Convex.exists_quantile_map_inner_cond_eq_add_of_volume_le). Averaging over the levels identifies this constant as ⟪u, v⟫, where v is the difference of the centroids of L and K (TauCeti.setIntegral_quantile_map_inner_cond). Since the quantiles sweep out the heights of interior points, interior L and v +ᵥ interior K have the same heights along every unit vector, so by separation of points from open convex sets they coincide (Convex.subset_of_image_inner_subset). Passing to closures gives L = v +ᵥ K. The additive equality case reduces to the multiplicative one by rescaling A and B to unit volume.

The hypothesis of positive volume is equivalent to nonempty interior (Convex.addHaar_pos_iff_interior_nonempty).

Main results #

References #

theorem Convex.exists_interior_eq_vadd_interior_of_volume_le {E : Type u_1} [NormedAddCommGroup E] [InnerProductSpace ℝ E] [FiniteDimensional ℝ E] [MeasurableSpace E] [BorelSpace E] {K L : Set E} (hK : Convex ℝ K) (hKb : Bornology.IsBounded K) (hK0 : MeasureTheory.volume K ≠ 0) (hL : Convex ℝ L) (hLb : Bornology.IsBounded L) (hL0 : MeasureTheory.volume L ≠ 0) {θ : ℝ} (hθ₀ : 0 < θ) (hθ₁ : θ < 1) (hvol : MeasureTheory.volume ((1 - θ) • K + θ • L) ≤ MeasureTheory.volume K ^ (1 - θ) * MeasureTheory.volume L ^ θ) :
∃ (v : E), interior L = v +ᵥ interior K

Translated interiors under Brunn–Minkowski equality. Let K and L be bounded convex sets of positive volume and 0 < θ < 1. If the volume of (1 - θ) • K + θ • L is at most volume K ^ (1 - θ) * volume L ^ θ, that is, if the multiplicative Brunn–Minkowski inequality is an equality, then the interior of L is a translate of the interior of K.

theorem Convex.exists_eq_vadd_of_volume_le {E : Type u_1} [NormedAddCommGroup E] [InnerProductSpace ℝ E] [FiniteDimensional ℝ E] [MeasurableSpace E] [BorelSpace E] {K L : Set E} (hK : Convex ℝ K) (hKc : IsCompact K) (hK0 : MeasureTheory.volume K ≠ 0) (hL : Convex ℝ L) (hLc : IsCompact L) (hL0 : MeasureTheory.volume L ≠ 0) {θ : ℝ} (hθ₀ : 0 < θ) (hθ₁ : θ < 1) (hvol : MeasureTheory.volume ((1 - θ) • K + θ • L) ≤ MeasureTheory.volume K ^ (1 - θ) * MeasureTheory.volume L ^ θ) :
∃ (v : E), L = v +ᵥ K

Brunn–Minkowski equality for convex bodies, multiplicative form. Let K and L be compact convex sets of positive volume and 0 < θ < 1. If the multiplicative Brunn–Minkowski inequality volume K ^ (1 - θ) * volume L ^ θ ≤ volume ((1 - θ) • K + θ • L) is an equality, then L is a translate of K.

theorem Convex.volume_smul_add_smul_eq_iff {E : Type u_1} [NormedAddCommGroup E] [InnerProductSpace ℝ E] [FiniteDimensional ℝ E] [MeasurableSpace E] [BorelSpace E] {K L : Set E} (hK : Convex ℝ K) (hKc : IsCompact K) (hK0 : MeasureTheory.volume K ≠ 0) (hL : Convex ℝ L) (hLc : IsCompact L) (hL0 : MeasureTheory.volume L ≠ 0) {θ : ℝ} (hθ₀ : 0 < θ) (hθ₁ : θ < 1) :
MeasureTheory.volume ((1 - θ) • K + θ • L) = MeasureTheory.volume K ^ (1 - θ) * MeasureTheory.volume L ^ θ ↔ ∃ (v : E), L = v +ᵥ K

Equality cases of multiplicative Brunn–Minkowski for convex bodies. Let K and L be compact convex sets of positive volume and 0 < θ < 1. Then volume ((1 - θ) • K + θ • L) = volume K ^ (1 - θ) * volume L ^ θ exactly when L is a translate of K.

theorem Convex.exists_eq_vadd_smul_of_volume_real_rpow_le {E : Type u_1} [NormedAddCommGroup E] [InnerProductSpace ℝ E] [FiniteDimensional ℝ E] [MeasurableSpace E] [BorelSpace E] {A B : Set E} (hA : Convex ℝ A) (hAc : IsCompact A) (hA0 : MeasureTheory.volume A ≠ 0) (hB : Convex ℝ B) (hBc : IsCompact B) (hB0 : MeasureTheory.volume B ≠ 0) {c d : ℝ} (hc : 0 < c) (hd : 0 < d) (h : MeasureTheory.volume.real (c • A + d • B) ^ (↑(Module.finrank ℝ E))⁻¹ ≤ c * MeasureTheory.volume.real A ^ (↑(Module.finrank ℝ E))⁻¹ + d * MeasureTheory.volume.real B ^ (↑(Module.finrank ℝ E))⁻¹) :
∃ (v : E) (r : ℝ), 0 < r ∧ B = v +ᵥ r • A

Brunn–Minkowski equality for convex bodies, additive form. Let A and B be compact convex sets of positive volume in a space of dimension n, and let c, d > 0. If the reverse bound volume (c • A + d • B) ^ (1 / n) ≤ c * volume A ^ (1 / n) + d * volume B ^ (1 / n) holds, then B is homothetic to A: B = v +ᵥ r • A for some vector v and some r > 0. In positive dimension, the bound gives equality in the Brunn–Minkowski inequality; in dimension zero, homothety follows from nonempty A and B alone.

theorem Convex.volume_real_smul_add_smul_rpow_eq_iff {E : Type u_1} [NormedAddCommGroup E] [InnerProductSpace ℝ E] [FiniteDimensional ℝ E] [MeasurableSpace E] [BorelSpace E] (hn : Module.finrank ℝ E ≠ 0) {A B : Set E} (hA : Convex ℝ A) (hAc : IsCompact A) (hA0 : MeasureTheory.volume A ≠ 0) (hB : Convex ℝ B) (hBc : IsCompact B) (hB0 : MeasureTheory.volume B ≠ 0) {c d : ℝ} (hc : 0 < c) (hd : 0 < d) :

Equality cases of Brunn–Minkowski for convex bodies. Let A and B be compact convex sets of positive volume in a space of positive dimension n, and let c, d > 0. Then volume (c • A + d • B) ^ (1 / n) = c * volume A ^ (1 / n) + d * volume B ^ (1 / n) exactly when B is homothetic to A, that is, B = v +ᵥ r • A for some vector v and some r > 0.