Skip to content
Menu
中文Sign in
← All questions

Why does the persistence of Gibbs overshoot distinguish pointwise convergence from uniform convergence for the square wave?

Pointwise convergence requires that for every fixed xx, SN(x)→f(x)S_N(x) \to f(x). This holds true except at the jump, where it converges to the midpoint. Uniform convergence requires that sup⁡x∣SN(x)−f(x)∣→0\sup_x |S_N(x) - f(x)| \to 0. Because the Gibbs overshoot maintains a non-zero peak error near the jump regardless of NN, the maximum difference never goes to zero. Thus, the convergence is pointwise but not uniform.

Conditions

  • Comparing definitions of pointwise and uniform convergence.
  • Observing the error bound near the discontinuity.

Reasoning, step by step

  1. Define pointwise convergence: limit of sequence at each point.
  2. Define uniform convergence: limit of the supremum of the error.
  3. Analyze the error of SN(x)S_N(x) vs the target function.
  4. Identify that the Gibbs peak creates a persistent lower bound on the sup-norm error.
  5. Conclude that since the max error doesn't vanish, convergence is not uniform.

Example

The script states: 'This distinguishes pointwise convergence from uniform convergence across the jumps.'

Common misconceptions

  • Thinking that pointwise convergence implies uniform convergence for continuous piecewise functions.
  • Ignoring the impact of isolated discontinuities on global convergence metrics.

Watch the explanation

0:19 – 0:30Watch this moment ↗

Explore next

Answers are generated from source material and independently checked. Consult the original video or creator if something is unclear.