Weierstrass function

The Weierstrass function is a famous mathematical object used in real analysis. It is a real-valued function that is continuous everywhere but differentiable nowhere. To be continuous means the graph is a single, unbroken path without any gaps. To be differentiable means you can find a specific slope or tangent line at any point. Most smooth curves become straight lines if you zoom in closely enough. However, the Weierstrass function remains jagged at every possible scale. This unique property makes it a classic example of a fractal curve.

.]

We can understand the mechanism of this function through its mathematical construction. Karl Weierstrass originally defined it using a Fourier series. This series involves adding up many different waves together. The formula uses a constant called 'a' and a positive odd integer called 'b'. The function stays continuous because the sum of the terms is bounded. Specifically, the Weierstrass M-test shows that the infinite series converges uniformly. This uniform convergence ensures the resulting path is continuous and even uniformly continuous.

This function is classified as a fractal because of its self-similarity. A fractal is a shape that looks similar no matter how much you zoom in. In a standard smooth function, zooming in eventually reveals a straight line. In the Weierstrass function, the curve is never monotone between any two points. This means it never stays strictly increasing or decreasing. Instead, it exhibits rapid oscillations that prevent a smooth slope from forming. Because of these constant fluctuations, the function has detail at every level of magnification.

Historically, the discovery of this function caused a massive shift in mathematics. Karl Weierstrass presented his findings to the Königliche Akademie der Wissenschaften on 18 July 1872. Before this, many prominent mathematicians, including Gauss, assumed that continuity implied differentiability almost everywhere. They believed that non-differentiable points would only occur at isolated spots. Weierstrass proved this intuition was wrong by showing a function could be jagged everywhere. This discovery overturned many existing proofs that relied on vague ideas of geometric smoothness.

Many of Weierstrass's contemporaries reacted to this discovery with frustration. The function was viewed as a "pathological" example, meaning it behaved in a way that seemed broken or unnatural. The mathematician Charles Hermite even described such functions as a "lamentable scourge." It was very difficult for people to visualize these shapes before the invention of computers. It was not until the next century that these jagged models gained wide acceptance. Today, they are essential for modeling Brownian motion, which describes the random movement of particles.

There are several related mathematical ideas that help explain the Weierstrass function's complexity. For instance, the function is Hölder continuous with an exponent of alpha. This describes how the function's values change relative to its input. However, the function is not Lipschitz continuous, which is a stricter type of smoothness. The function is also related to the earlier Riemann function. While Riemann claimed his function was nowhere differentiable, Weierstrass noted he found no published evidence of this. Later, in 1916, G. H. Hardy provided a rigorous confirmation regarding the function's derivatives.

Finally, the Weierstrass function is not just a strange exception; it is actually quite typical. In a topological sense, the set of nowhere-differentiable functions is "comeager" in the space of continuous functions. This means that, mathematically speaking, most continuous functions actually behave like the Weierstrass function. When using the Wiener measure, the collection of functions that are differentiable at even a single point has a measure of zero. This reveals that the smooth lines we study in basic calculus are actually the rare exceptions in the vast world of continuous mathematics.