Documentation

TauCeti.Analysis.CompletelyMonotone.FiniteMixture

Finite mixtures of completely monotone exponential functions #

The extreme rays of the cone of completely monotone functions are the exponential functions t ↦ exp (-p * t) with p ≥ 0. This file develops their finite positive mixtures. It packages a finite mixture both as an explicit sum and as the Laplace transform of the corresponding finite sum of weighted Dirac measures.

This is the finite-support case of the measure representation in Bernstein's theorem. In particular, a single atom of unit weight at p = 1 represents t ↦ exp (-t), one of the acceptance examples in the one-parameter-semigroups roadmap.

Main declarations #

References #

noncomputable def TauCeti.finiteExponentialMixture {ι : Type u_1} (s : Finset ι) (w p : ι → NNReal) (t : ℝ) :

A finite positive mixture of the exponential extreme rays t ↦ exp (-p * t).

The weights and rates are indexed by a finset. Both take values in ℝ≥0, making their nonnegativity intrinsic to the data.

Equations
Instances For
    @[simp]

    The empty exponential mixture is the zero function.

    @[simp]
    theorem TauCeti.finiteExponentialMixture_singleton {ι : Type u_1} (i : ι) (w p : ι → NNReal) (t : ℝ) :
    finiteExponentialMixture {i} w p t = ↑(w i) * Real.exp (-↑(p i) * t)

    A singleton exponential mixture is one weighted exponential extreme ray.

    Every finite positive mixture of exponential extreme rays is completely monotone.

    noncomputable def TauCeti.finiteExponentialMixtureMeasure {ι : Type u_1} (s : Finset ι) (w p : ι → NNReal) :

    The finite positive measure that places weight w i at each rate p i.

    Repeated rates are intentionally allowed: measure addition combines their weights.

    Equations
    Instances For
      @[simp]
      theorem TauCeti.finiteExponentialMixtureMeasure_apply {ι : Type u_1} (s : Finset ι) (w p : ι → NNReal) {A : Set NNReal} (hA : MeasurableSet A) :
      (finiteExponentialMixtureMeasure s w p) A = ∑ i ∈ s, A.indicator (fun (x : NNReal) => ↑(w i)) (p i)

      The representing measure of a measurable set is the sum of the weights at rates in the set.

      @[simp]

      The empty exponential mixture has the zero representing measure.

      @[simp]

      A singleton mixture has the corresponding weighted Dirac representing measure.

      A finite exponential mixture is the Laplace transform of its discrete representing measure.

      The Laplace transform of the unit Dirac mass at 1 is t ↦ exp (-t).

      This is the discrete representing-measure acceptance example for Bernstein's theorem.

      The exponential t ↦ exp (-t) is the unit-weight, unit-rate finite exponential mixture.