NRR.Multivalued.Separator.ToNiceMV — separators as nice multivalued functions #
Every top–bottom separator S over a nonempty metric base X gives a nice multivalued function
S.toNiceMV, whose scalar observable is the signed distance to the carrier. Its zero set is
exactly the carrier: the graph of the separating multifunction is recovered as the zero set of a
single continuous scalar function, with the required strict endpoint signs coming from the signed
distance being negative at the bottom boundary and positive at the top boundary.
Separators also pull back along a continuous base map f : Y → X through the product map
(y, t) ↦ (f y, t): carrier, lower, and upper regions are pulled back by preimage, which preserves
closedness, openness, disjointness, complements, and the boundary placement. The pulled-back
separator's nice multivalued function has the same zero set as the pullback of the original nice
multivalued function, even though the two signed distances are computed in different product metrics
and need not agree pointwise.
The nice multivalued function attached to a top–bottom separator: its scalar observable is the signed distance to the carrier, so its zero set is exactly the carrier.
Equations
- S.toNiceMV = { evalMap := S.signedDistanceMap, left_neg := ⋯, right_pos := ⋯ }
Instances For
Pull back a top–bottom separator S on X along a continuous base map f : Y → X through the
product map (y, t) ↦ (f y, t). Carrier, lower, and upper regions are pulled back by preimage.
Equations
- One or more equations did not get rendered due to their size.