Documentation

Mathlib.Analysis.Convex.MetricSpace

Convex metric spaces #

A convex metric space is a convex space with a compatible metric structure. Concretely, we ask for dist(∑ tᵢ xᵢ, ∑ tᵢ yᵢ) ≤ ∑ tᵢ dist(xᵢ, yᵢ), which is what one would expect from the triangle inequality.

Main results #

TODO #

A convex metric space is a real convex space with a compatible metric structure. Concretely, we ask for dist(∑ tᵢ xᵢ, ∑ tᵢ yᵢ) ≤ ∑ tᵢ dist(xᵢ, yᵢ), which is what one would expect from the triangle inequality.

In particular, convex subsets of normed affine spaces are convex metric spaces.

Instances
    theorem dist_convexCombination_map_le {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {ι : Type u_2} (f : StdSimplex ι) (x y : ιX) :

    dist(∑ tᵢ xᵢ, ∑ tᵢ yᵢ) ≤ ∑ tᵢ dist(xᵢ, yᵢ)

    theorem dist_convexCombination_left_le {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] (f : StdSimplex X) (x : X) :
    dist (ConvexSpace.convexCombination f) x f.sum fun (i : X) (r : ) => r * dist i x
    @[simp]
    theorem dist_convexComboPair_left {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {s t : } (hs : 0 s) (ht : 0 t) (h : s + t = 1) (x y : X) :
    dist (convexComboPair s t hs ht h x y) x = t * dist x y
    @[simp]
    theorem dist_convexComboPair_right {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {s t : } (hs : 0 s) (ht : 0 t) (h : s + t = 1) (x y : X) :
    dist (convexComboPair s t hs ht h x y) y = s * dist x y
    @[simp]
    theorem dist_left_convexComboPair {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {s t : } (hs : 0 s) (ht : 0 t) (h : s + t = 1) (x y : X) :
    dist x (convexComboPair s t hs ht h x y) = t * dist x y
    @[simp]
    theorem dist_right_convexComboPair {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {s t : } (hs : 0 s) (ht : 0 t) (h : s + t = 1) (x y : X) :
    dist y (convexComboPair s t hs ht h x y) = s * dist x y
    theorem dist_convexComboPair_convexComboPair {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {s t s' t' : } (hs : 0 s) (ht : 0 t) (h : s + t = 1) (hs' : 0 s') (ht' : 0 t') (h' : s' + t' = 1) (x y : X) :
    dist (convexComboPair s t hs ht h x y) (convexComboPair s' t' hs' ht' h' x y) = |s - s'| * dist x y

    dist(sx + (1-s)y, s'x + (1-s')y) = |s - s'| dist(x, y).

    See dist_convexComboPair_convexComboPair_le for the version where the weights are fixed and the points change.

    theorem dist_convexComboPair_convexComboPair_le {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {s t : } (hs : 0 s) (ht : 0 t) (h : s + t = 1) (x y x' y' : X) :
    dist (convexComboPair s t hs ht h x y) (convexComboPair s t hs ht h x' y') s * dist x x' + t * dist y y'

    dist(sx + (1-s)y, sx' + (1-s)y') ≤ s dist(x, x') + (1-s) dist(y, y').

    See dist_convexComboPair_convexComboPair for the version where the points are fixed and the weights change.

    theorem continuous_convexComboPair {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] :
    Continuous fun (x : (Set.Icc 0 1) × X × X) => convexComboPair (↑x.1) (1 - x.1) x.2.1 x.2.2

    The convex combination (t, p, q) ↦ t • p + (1 - t) • q is continuous on [0, 1] × X × X for a convex metric space X.

    theorem continuous_convexComboPair_of_isBounded {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] {T : Type u_2} [TopologicalSpace T] (f : T) (hf : Continuous f) (hf0 : ∀ (t : T), 0 f t) (hf1 : ∀ (t : T), f t 1) (x y : TX) (hx : ContinuousOn x (f ⁻¹' {0})) (hy : ContinuousOn y (f ⁻¹' {1})) (hx' : Bornology.IsBounded (Set.range x)) (hy' : Bornology.IsBounded (Set.range y)) :
    Continuous fun (i : T) => convexComboPair (f i) (1 - f i) (x i) (y i)
    theorem continuous_convexComboPair' {X : Type u_1} [ConvexSpace X] [MetricSpace X] [IsConvexMetricSpace X] [BoundedSpace X] {T : Type u_2} [TopologicalSpace T] (f : T) (hf : Continuous f) (hf0 : ∀ (t : T), 0 f t) (hf1 : ∀ (t : T), f t 1) (x y : TX) (hx : ContinuousOn x (f ⁻¹' {0})) (hy : ContinuousOn y (f ⁻¹' {1})) :
    Continuous fun (i : T) => convexComboPair (f i) (1 - f i) (x i) (y i)

    When X is a bounded convex metric space, to check continuity of t ↦ f(t) • x(t) + (1 - f(t)) • y(t) it suffices to show that f is continuous, x is continuous away from f(t) = 0, and y is continuous away from f(t) = 1.

    @[implicit_reducible]
    noncomputable def ConvexSpace.ofConvex {R : Type u_2} {E : Type u_3} [LinearOrder R] [Field R] [IsStrictOrderedRing R] [AddCommGroup E] [Module R E] {S : Set E} (H : Convex R S) :

    A convex subset of a vector space is a convex space.

    Equations
    Instances For