logo

Proof of the Inverse Function Theorem in Complex Analysis 📂Complex Anaylsis

Proof of the Inverse Function Theorem in Complex Analysis

Theorem 1

A function f:CCf : \mathbb{C} \to \mathbb{C} that is analytic at α\alpha and satisfies f(α)0f ' (\alpha) \ne 0 exists in the region N(f(α))\mathcal{N} \left( f(\alpha) \right), where f1f^{-1} exists.

Explanation

Consider the condition given by f(α)0f ' (\alpha) \ne 0.

When thought of as a real function, it implies the function is either increasing or decreasing, which is a condition for the existence of an inverse function. Geometrically, this refers to a Smooth function, indicating there are no abrupt changes in direction or bends. What must be noted in the inverse function theorem is that even if such conditions are met, it doesn’t mean the inverse function itself exists universally but rather within a local limit.

Proof

It suffices to show that equation w=f(z)w = f(z) has a unique solution for wN(f(α))w \in \mathcal{N} (f(\alpha)).


If we set β:=f(α)g(z):=f(z)β \beta := f(\alpha) \\ g(z) := f(z) - \beta , then g(α)=0g(\alpha) = 0 and g(α)0g ' (\alpha) \ne 0. This means that α\alpha is a simple zero of gg, and there exists ρ>0\rho >0 that satisfies g(z)0g(z) \ne 0 in zαρ|z - \alpha | \le \rho.

For circle C:zα=ρ\mathscr{C}: |z - \alpha| = \rho, define h(z):=γh(z) := -\gamma that satisfies m:=minCg(z)γ<m m := \min_{\mathscr{C}} |g(z)| \\ |\gamma| < m . Then, the following holds in C\mathscr{C}: g(z)0h(z)=γ=γ<mg(z) g(z) \ne 0 \\ |h(z) | = | - \gamma | = |\gamma| < m \le |g(z)|

Rouché’s Theorem: If gg and hh are analytic on the simple closed path C\mathscr{C} and satisfy h(z)g(z)|h(z)| \le |g(z)| on C\mathscr{C}, then gg and g+hg + h have the same number of zeros inside C\mathscr{C}.

By Rouché’s Theorem, gg and g+h=gγg + h = g - \gamma have the same number of zeros inside C\mathscr{C}.

However, as seen before, since gg has only a single simple zero α\alpha, the zero satisfying g(z)γ=0g(z) - \gamma = 0 is also unique inside C\mathscr{C}. Therefore, it can be concluded that equation g(z)=γg(z) = \gamma has a unique solution inside C\mathscr{C}.

If we now let w=β+γw = \beta + \gamma, f(z)β=wβ f(z) - \beta = w - \beta it means that w=f(z)w = f(z) has a unique solution inside C\mathscr{C} at N(α):zα<ρ\mathcal{N}(\alpha): |z - \alpha| < \rho.

See Also


  1. Osborne (1999). Complex variables and their applications: p193. ↩︎