Banach Fixed-Point Theorem [dark version]

Q: Why does completeness matter in Banach’s theorem?

Completeness is what turns “Cauchy” into “convergent.” After defining xn+1 = F(xn), the proof derives bounds that make D(xn, xm) arbitrarily small for large indices, so (xn) is Cauchy. Completeness of (X, D) then guarantees there is a limit x* in X. Without completeness, a Cauchy sequence might converge only in a larger space, not inside X, and the fixed point could fail to exist in X.

Q: How does the contraction inequality produce a geometric decay like Q^n?

The contraction gives D(xn+1, xn) = D(F(xn), F(xn−1)) ≤ Q·D(xn, xn−1). Applying the same inequality repeatedly yields D(xn+1, xn) ≤ Q^n·D(x1, x0). This geometric factor is crucial: because Q < 1, powers of Q shrink rapidly, making later terms close together.

Q: Why isn’t controlling only neighboring distances enough to prove the sequence is Cauchy?

A Cauchy sequence requires D(xn, xm) to be small for all m < n, not just for m = n−1. The proof uses the triangle inequality to bridge the gap: D(xn, xm) ≤ D(xn, xn−1) + D(xn−1, xn−2) + … + D(xm+1, xm). Each summand has its own Q-power bound, and the total becomes a geometric series.

Q: What role does the triangle inequality play in the Cauchy estimate?

It allows the proof to “walk” from xn back to xm in steps of one iteration. Since D(xn, xm) is no larger than the sum of the step distances, and each step distance is bounded by a power of Q, the entire distance is bounded by a finite geometric sum. That sum stays controlled because Q is strictly less than 1.

Q: How does the proof show the limit x* is a fixed point?

Once xn → x*, continuity is used to pass the limit through F: F(x*) = F(lim xn) = lim F(xn). Because F(xn) = xn+1 by construction, lim F(xn) = lim xn+1 = x*. So x* satisfies F(x*) = x*.

Q: Why can’t there be two different fixed points under a contraction?

Assume x* and x̂ are fixed points with x* ≠ x̂. Then D(x*, x̂) = D(F(x*), F(x̂)) ≤ Q·D(x*, x̂). Since x* ≠ x̂ implies D(x*, x̂) > 0, dividing both sides by D(x*, x̂) yields 1 ≤ Q, contradicting Q < 1. Hence the fixed point is unique.

TL;DR

Banach’s theorem requires a complete metric space (X, D) and a contraction F: X → X with a constant Q satisfying 0 ≤ Q < 1.

Briefing Cornell Notes

Briefing

Banach’s Fixed-Point Theorem guarantees a unique fixed point for a contraction on a complete metric space—and it also provides a practical way to approximate that point. The setup is precise: start with a complete metric space (X, D) and a function F: X → X that shrinks distances by a uniform factor Q, where 0 ≤ Q < 1. Concretely, for every pair x, y in X, the distance between their images satisfies D(F(x), F(y)) ≤ Q·D(x, y). Under these conditions, there exists exactly one point x* in X such that F(x*) = x*, and iterating F from any starting point converges to x*.

The proof begins by constructing an iterative sequence. Pick an arbitrary starting value x0 in X and define xn+1 = F(xn). To show the sequence converges, it’s enough (because X is complete) to prove it is Cauchy. The contraction property yields a key estimate for successive terms: D(xn+1, xn) = D(F(xn), F(xn−1)) ≤ Q·D(xn, xn−1). Repeating the same reasoning step-by-step gives a geometric bound on neighbor distances: D(xn+1, xn) ≤ Q^n·D(x1, x0).

Neighbor control alone doesn’t yet guarantee the Cauchy property, so the argument uses the triangle inequality to relate distant terms. For m < n, D(xn, xm) is bounded by summing the distances along the “path” xn → xn−1 → … → xm. Each link in that chain inherits a power of Q from the neighbor estimate, producing a geometric series. Because Q is strictly less than 1, the series has a finite bound, and the overall distance D(xn, xm) can be made arbitrarily small by choosing m large enough. That establishes that (xn) is Cauchy.

Completeness then forces convergence: there is some limit x* in X with xn → x*. The next step is to show x* is actually a fixed point. Contractions are continuous in metric spaces, so taking limits through F is valid: F(x*) = F(lim xn) = lim F(xn) = lim xn+1 = x*. Finally, uniqueness follows from a short contradiction argument. If two fixed points x* and x̂ existed with x* ≠ x̂, then applying the contraction inequality to them would give D(x*, x̂) = D(F(x*), F(x̂)) ≤ Q·D(x*, x̂). Since D(x*, x̂) > 0 when the points differ, dividing would imply 1 ≤ Q, contradicting Q < 1. Therefore, only one fixed point can exist.

The theorem matters because it turns an abstract existence-and-uniqueness claim into a workable algorithm: repeated iteration of F from any starting point converges to the same x*. This is why the result is widely used in analysis, including proving uniqueness of solutions to differential equations under suitable conditions.

Cornell Notes

Banach’s Fixed-Point Theorem applies to a complete metric space (X, D) and a contraction F: X → X with a constant Q satisfying 0 ≤ Q < 1. The contraction condition D(F(x), F(y)) ≤ Q·D(x, y) forces the iterates xn+1 = F(xn) to form a Cauchy sequence. Completeness then guarantees convergence to some limit x*. Continuity of contractions lets the limit pass through F, yielding F(x*) = x*, so x* is a fixed point. A final inequality shows fixed points cannot be distinct: if two fixed points existed, the contraction would imply 1 ≤ Q, contradicting Q < 1. The theorem also provides an iteration method that converges to the unique fixed point from any starting point.

Why does completeness matter in Banach’s theorem?

Completeness is what turns “Cauchy” into “convergent.” After defining xn+1 = F(xn), the proof derives bounds that make D(xn, xm) arbitrarily small for large indices, so (xn) is Cauchy. Completeness of (X, D) then guarantees there is a limit x* in X. Without completeness, a Cauchy sequence might converge only in a larger space, not inside X, and the fixed point could fail to exist in X.

How does the contraction inequality produce a geometric decay like Q^n?

The contraction gives D(xn+1, xn) = D(F(xn), F(xn−1)) ≤ Q·D(xn, xn−1). Applying the same inequality repeatedly yields D(xn+1, xn) ≤ Q^n·D(x1, x0). This geometric factor is crucial: because Q < 1, powers of Q shrink rapidly, making later terms close together.

Why isn’t controlling only neighboring distances enough to prove the sequence is Cauchy?

A Cauchy sequence requires D(xn, xm) to be small for all m < n, not just for m = n−1. The proof uses the triangle inequality to bridge the gap: D(xn, xm) ≤ D(xn, xn−1) + D(xn−1, xn−2) + … + D(xm+1, xm). Each summand has its own Q-power bound, and the total becomes a geometric series.

What role does the triangle inequality play in the Cauchy estimate?

It allows the proof to “walk” from xn back to xm in steps of one iteration. Since D(xn, xm) is no larger than the sum of the step distances, and each step distance is bounded by a power of Q, the entire distance is bounded by a finite geometric sum. That sum stays controlled because Q is strictly less than 1.

How does the proof show the limit x* is a fixed point?