LF2 Projective Weights #
Category: 3-Local (LF2 measurable partitions on the projective target and their weight normalisation).
Spec §4:
MeasurablePartition— measurable partition ofPup to null sets (spec §4.1).projectiveWeight— the weight of an outcome region under the pushforward of a preparation measure (spec §4.2).weights_sum_eq_one— normalisation (spec §4.3).
The partition is stated relative to the preparation's pushforward measure
π*μprep. The measure bridges concern the sector's reference measure μL;
they do not identify π*μprep with μFS for an arbitrary preparation measure.
A μFS-relative partition would also suffice if π*μprep ≪ μFS, since absolute
continuity transfers its null overlaps and null uncovered set to π*μprep.
No such relationship is assumed here. See specs/LF2-plan.md §2.3 for the
partition interface.
weights_sum_eq_one uses measurability of π, the probability preparation
measure and the partition hypotheses. It does not use a Fubini–Study measure
or the sector's symmetry and measure-bridge properties.
A finite measurable partition of P up to μ-null sets: measurable
parts, pairwise intersections of μ-measure zero, and complement of the
union of μ-measure zero.
The parts, indexed by
Fin n.- measurable (i : Fin n) : MeasurableSet (self.parts i)
Each part is measurable.
Pairwise intersections are null.
The complement of the union is null.
Instances For
Projective weight of an outcome region O ⊆ P under the pushforward of
a preparation measure. This definition allows arbitrary measures and sets;
probability normalisation and measurable parts are assumed in weights_sum_eq_one.
Equations
- CSD.LF2.projectiveWeight D μprep O = (MeasureTheory.Measure.map D.π μprep) O
Instances For
Unfolding lemma: projectiveWeight is the pushforward measure of the
region. Not @[simp] — leaving projectiveWeight as a named
abstraction makes downstream proofs and theorem statements
readable; callers unfold projectiveWeight (or rewrite with this
lemma explicitly) when they need to expose the underlying
pushforward.
Spec §4.3 normalisation. For a probability preparation measure and a
measurable partition of P up to π*μprep-null sets, the weights sum
to one.