Smooth positive ratio of cooriented regular defining functions
ProvedDifferentialTopology.positive_ratio_of_cooriented_regular_defining_functionscontact-topologydifferential-topologyreal-analysissmooth-functions
Let be smooth. Suppose zero is a regular value of each function, their zero sets coincide, and they have the same sign away from that set:
Then there is a unique smooth positive function such that
Thus the ordinary quotient , initially defined off the common zero curve, extends smoothly and positively across it. The conclusion also covers an empty zero set.
This is a smooth division consequence of Hadamard's lemma and the inverse function theorem. In contact geometry it permits cooriented regular dividing functions to be normalized to a common normal coefficient without changing the corresponding contact plane field under positive rescaling.
Preamble
import Mathlib open scoped ContDiff
Formal statement
theorem DifferentialTopology.positive_ratio_of_cooriented_regular_defining_functions
(u v : (ℝ × ℝ) → ℝ) (hu : ContDiff ℝ ∞ u) (hv : ContDiff ℝ ∞ v)
(hregu : ∀ x, u x = 0 → fderiv ℝ u x ≠ 0)
(hregv : ∀ x, v x = 0 → fderiv ℝ v x ≠ 0)
(hzero : ∀ x, u x = 0 ↔ v x = 0)
(hsign : ∀ x, v x ≠ 0 → 0 < u x*v x) :
∃! a : (ℝ × ℝ) → ℝ, ContDiff ℝ ∞ a ∧ (∀ x, 0 < a x) ∧ ∀ x, u x = a x*v x := by sorrySource
Smooth division corollary of Hadamard's lemma with parameters and the inverse function theorem: J. Milnor, Morse Theory (1963), printed p.5, Lemma 2.1, https://www.maths.ed.ac.uk/~v1ranick/papers/milnmors.pdf. The global positive-factor conclusion is the stated corollary, not the verbatim lemma. Used for invariant contact-form normalization in Etnyre, Convex Surfaces in Contact Geometry notes, printed pp.11–13, proof of Theorem 2.26.