Method of exhaustion #
If μ, ν
are two measures with ν
s-finite, then there exists a set s
such that
μ
is sigma-finite on s
, and for all sets t ⊆ sᶜ
, either ν t = 0
or μ t = ∞
.
Main definitions #
MeasureTheory.Measure.sigmaFiniteSetWRT
: if such a set exists,μ.sigmaFiniteSetWRT ν
is a measurable set such thatμ.restrict (μ.sigmaFiniteSetWRT ν)
is sigma-finite and for all setst ⊆ (μ.sigmaFiniteSetWRT ν)ᶜ
, eitherν t = 0
orμ t = ∞
. If no such set exists (which is only possible ifν
is not s-finite), we defineμ.sigmaFiniteSetWRT ν = ∅
.MeasureTheory.Measure.sigmaFiniteSet
: for an s-finite measureμ
, a measurable set such thatμ.restrict μ.sigmaFiniteSet
is sigma-finite, and for all setss ⊆ μ.sigmaFiniteSetᶜ
, eitherμ s = 0
orμ s = ∞
. Defined asμ.sigmaFiniteSetWRT μ
.
Main statements #
measure_eq_top_of_subset_compl_sigmaFiniteSetWRT
: for s-finiteν
, for all setss
in(sigmaFiniteSetWRT μ ν)ᶜ
, ifν s ≠ 0
thenμ s = ∞
.An instance showing that
μ.restrict (sigmaFiniteSetWRT μ ν)
is sigma-finite.restrict_compl_sigmaFiniteSetWRT
: ifμ ≪ ν
andν
is s-finite, thenμ.restrict (μ.sigmaFiniteSetWRT ν)ᶜ = ∞ • ν.restrict (μ.sigmaFiniteSetWRT ν)ᶜ
. As a consequence, that restriction is s-finite.An instance showing that
μ.restrict μ.sigmaFiniteSet
is sigma-finite.restrict_compl_sigmaFiniteSet_eq_zero_or_top
: the measureμ.restrict μ.sigmaFiniteSetᶜ
takes only two values: 0 and ∞ .measure_compl_sigmaFiniteSet_eq_zero_iff_sigmaFinite
: a measureμ
is sigma-finite iffμ μ.sigmaFiniteSetᶜ = 0
.
References #
- [P. R. Halmos, Measure theory, 17.3 and 30.11][halmos1950measure]
A measurable set such that μ.restrict (μ.sigmaFiniteSetWRT ν)
is sigma-finite and for all
measurable sets t ⊆ sᶜ
, either ν t = 0
or μ t = ∞
.
Equations
Instances For
Equations
- ⋯ = ⋯
We prove that the condition in the definition of sigmaFiniteSetWRT
is true for finite
measures. Since every s-finite measure is absolutely continuous with respect to a finite measure,
the condition will then also be true for s-finite measures.
Let C
be the supremum of ν s
over all measurable sets s
such that μ.restrict s
is
sigma-finite. C
is finite since ν
is a finite measure. Then there exists a measurable set t
with μ.restrict t
sigma-finite such that ν t ≥ C - 1/n
.
A measurable set such that μ.restrict (μ.sigmaFiniteSetGE ν n)
is sigma-finite and
for C
the supremum of ν s
over all measurable sets s
with μ.restrict s
sigma-finite,
ν (μ.sigmaFiniteSetGE ν n) ≥ C - 1/n
.
Equations
- μ.sigmaFiniteSetGE ν n = ⋯.choose
Instances For
A measurable set such that μ.restrict (μ.sigmaFiniteSetWRT' ν)
is sigma-finite and
ν (μ.sigmaFiniteSetWRT' ν)
has maximal measure among such sets.
Instances For
μ.sigmaFiniteSetWRT' ν
has maximal ν
-measure among all measurable sets s
with sigma-finite
μ.restrict s
.
Auxiliary lemma for measure_eq_top_of_subset_compl_sigmaFiniteSetWRT'
.
For all sets s
in (μ.sigmaFiniteSetWRT ν)ᶜ
, if ν s ≠ 0
then μ s = ∞
.
For all sets s
in (μ.sigmaFiniteSetWRT ν)ᶜ
, if ν s ≠ 0
then μ s = ∞
.
A measurable set such that μ.restrict μ.sigmaFiniteSet
is sigma-finite,
and for all measurable sets s ⊆ μ.sigmaFiniteSetᶜ
, either μ s = 0
or μ s = ∞
.
Equations
- μ.sigmaFiniteSet = μ.sigmaFiniteSetWRT μ
Instances For
The measure μ.restrict μ.sigmaFiniteSetᶜ
takes only two values: 0 and ∞ .
The restriction of an s-finite measure μ
to μ.sigmaFiniteSet
is sigma-finite.
Equations
- ⋯ = ⋯
An s-finite measure μ
is sigma-finite iff μ μ.sigmaFiniteSetᶜ = 0
.