I was playing around with something this afternoon and stumbled on something like Gibbs phenomena or Runge phenomena for the Central Limit Theorem.
The first place most people encounter Gibbs phenomenon is in Fourier series for a step function. The Fourier series develops “bat ears” near the discontinuity. Here’s an example I blogged about before not with Fourier series but with analogous Chebyshev series.
The series converges rapidly in the middle of the flat parts, but under-shoots and over-shoots near the jumps in the step function.
Runge phenomena is similar, where interpolating functions under- and over-shoot the function they’re approximating.
Both plots above come from this post.
Here’s the example I ran across with the central limit theorem. The distribution of the average of a set of exponential random variables converges to the distribution of a normal random variable. The nice thing about the exponential distribution is that the averages have a familiar distribution: a gamma distribution. If each exponential has mean 1, the average has a gamma distribution with shape N and scale 1/N. The central limit theorem says this is converging in distribution to a normal distribution with the same mean and variance.
The plot below shows the difference between the density function of the average of N exponential random variables and the density function for its normal approximation, for N = 10 and for N = 400.
Notice that the orange line, corresponding to N = 400, is very flat, most of the time. That is, the normal approximation fits very well. But there’s this spike in the middle, something reminiscent of Gibbs phenomena or Runge phenomena. Going from 10 to 400 samples the average error decreases quite a bit, but the maximum error doesn’t go down much at all.
If you go back and look at the Central Limit Theorem, or its more quantitative counterpart the Berry-Esseen theorem, you’ll notice that it applies to the distribution function, not the density, or in other words, the CDF, not the PDF. I think the density functions do converge in this case because the exponential function has a smooth density, but the rate of convergence depends on the norm. It looks like the convergence is fast in square (L²) norm, but slow in sup norm. A little experiment shows that this is indeed the case.
Maybe the max norm doesn’t converge at all, i.e. the densities don’t converge pointwise. It looks like the max norm may be headed toward a horizontal asymptote, just like Gibbs phenomena.
Update: It seems we do not have uniform convergence. If we let N = 1,000,000, the sup norm of the error 0.1836. It appears the sup norm of the error is approaching a lower bound of approximately this value.
I stumbled on your post because I’m interested in the same questions.
In fact, assuming that the maximum of the error is at x=1+y with a small y, one finds that as N goes to infinity, y=sqrt(2/5N) and the maximum of the error is 13/(15 (e^(1/5) sqrt(5/pi ))) ~ 0.179033, independent of N.
I am looking for some references on this phenomenon, any hints?
Best