Contents
In this file we work with contents. A content λ
is a function from a certain class of subsets
(such as the the compact subsets) to ennreal
(or nnreal
) that is
- additive: If
K₁
andK₂
are disjoint sets in the domain ofλ
, thenλ(K₁ ∪ K₂) = λ(K₁) + λ(K₂)
; - subadditive: If
K₁
andK₂
are in the domain ofλ
, thenλ(K₁ ∪ K₂) ≤ λ(K₁) + λ(K₂)
; - monotone: If
K₁ ⊆ K₂
are in the domain ofλ
, thenλ(K₁) ≤ λ(K₂)
.
We show that:
- Given a content
λ
on compact sets, we get a countably subadditive map that vanishes at∅
. In Halmos (1950) this is called the inner contentλ*
ofλ
. - Given an inner content, we define an outer measure.
We don't explicitly define the type of contents. In this file we only work on contents on compact sets, and inner contents on open sets, and both contents and inner contents map into the extended nonnegative reals. However, in other applications other choices can be made, and it is not a priori clear what the best interface should be.
Main definitions
measure_theory.inner_content
: define an inner content from an contentmeasure_theory.outer_measure.of_content
: construct an outer measure from a content
References
- Paul Halmos (1950), Measure Theory, §53
- https://en.wikipedia.org/wiki/Content_(measure_theory)
Constructing the inner content of a content. From a content defined on the compact sets, we obtain a function defined on all open sets, by taking the supremum of the content of all compact subsets.
Equations
- measure_theory.inner_content μ U = ⨆ (K : topological_space.compacts G) (h : K.val ⊆ ↑U), μ K
This is "unbundled", because that it required for the API of induced_outer_measure
.
The inner content of a surpremum of opens is at most the sum of the individual inner contents.
The inner content of a union of sets is at most the sum of the individual inner contents.
This is the "unbundled" version of inner_content_Sup_nat
.
It required for the API of induced_outer_measure
.
Extending a content on compact sets to an outer measure on all sets.
Equations
- measure_theory.outer_measure.of_content μ h1 = measure_theory.induced_outer_measure (λ (U : set G) (hU : is_open U), measure_theory.inner_content μ ⟨U, hU⟩) is_open_empty _