Injective pieces of a differentiable map #
A measurable set on which a map has injective derivative admits a countable measurable partition on which the map is bi-Lipschitz. This decomposition permits summation of injective change-of-variables and area formulas with multiplicity, without requiring the original map to be injective.
The construction uses Mathlib's
exists_partition_approximatesLinearOn_of_hasFDerivWithinAt, with an absolute error scaled to
each candidate linear map, so that on every piece the map is within any prescribed relative error
δ of an injective linear map
(TauCeti.exists_partition_linear_approximation_of_hasFDerivWithinAt).
For δ < 1 such a piece has a Lipschitz inverse
(TauCeti.exists_antilipschitzWith_domRestrict_of_approx), which gives the bi-Lipschitz partition
TauCeti.exists_partition_lipschitz_antilipschitz_of_hasFDerivWithinAt.
References #
- L. C. Evans, R. F. Gariepy, Measure Theory and Fine Properties of Functions, revised edition, §3.3 (decomposition into injective pieces in the area formula).
A map within relative error δ of a linear map A on p satisfies
(1 - δ) ‖A x - A y‖ ≤ ‖f x - f y‖ there.
A map within relative error δ of a linear map A on p satisfies
‖f x - f y‖ ≤ (1 + δ) ‖A x - A y‖ there.
A map f within relative error ε < 1 of an injective linear map A on p, read through
A, is (1 + ε)-Lipschitz on A '' p with (1 - ε)⁻¹-Lipschitz inverse.
A map within relative error δ < 1 of an injective linear map on p has a Lipschitz inverse
on p.
A measurable set on which the derivative is injective has, for every δ > 0, a countable
measurable partition into pieces p n, each with a linear map A n that is injective whenever
the piece is nonempty and satisfies ‖f x - f y - A n (x - y)‖ ≤ δ ‖A n (x - y)‖ on the piece.
No global Lipschitz assumption is needed.
A measurable set on which the derivative is injective has a countable measurable partition into bi-Lipschitz pieces. No global Lipschitz assumption is needed.