LeanAide GPT-4 documentation

Module: Mathlib.Analysis.Calculus.InverseFunctionTheorem.ContDiff


ContDiffAt.to_localInverse

theorem ContDiffAt.to_localInverse : โˆ€ {๐•‚ : Type u_1} [inst : RCLike ๐•‚] {E : Type u_2} [inst_1 : NormedAddCommGroup E] [inst_2 : NormedSpace ๐•‚ E] {F : Type u_3} [inst_3 : NormedAddCommGroup F] [inst_4 : NormedSpace ๐•‚ F] [inst_5 : CompleteSpace E] {f : E โ†’ F} {f' : E โ‰ƒL[๐•‚] F} {a : E} {n : โ„•โˆž} (hf : ContDiffAt ๐•‚ n f a) (hf' : HasFDerivAt f (โ†‘f') a) (hn : 1 โ‰ค n), ContDiffAt ๐•‚ n (hf.localInverse hf' hn) (f a)

This theorem states that if you have a continuously differentiable function `f` from `E` to `F` over a ring `๐•‚` (which could be the real or complex numbers), that has an invertible derivative at a point `a`, then the inverse function of `f` is also continuously differentiable at the point `f(a)`. This inverse function is obtained by applying the `ContDiff.toPartialHomeomorph` function on `f`. The theorem requires the derivative of `f` at `a` to be given by the linear map `f'`, and the differentiability of `f` is specified to a degree `n`, where `n` is a natural number or infinity. Furthermore, `n` should be at least 1. The space `E` is assumed to be complete, which means that every Cauchy sequence in `E` converges to a limit that is also in `E`.

More concisely: If `f: E โ†’ F` is a continuously differentiable function with an invertible derivative at `a โˆˆ E` over a complete ring `๐•‚`, then the inverse function `g` of `f` is also continuously differentiable at `b = f(a)` with derivative `g' = (f')โปยน`.