Sign up ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

As we know, for $1<p<\infty$, the Fourier series of $f\in L^{p}(T)$ converges to $f$ in $L^{p}$-norm. But is there any results concerning the convergence of Fourier series in $L^{\infty}$-norm? Since $L^{\infty}(T)$ is not separable, the trigonometric system fails to form a Schauder basis of $L^{\infty}(T)$, this implies that the Fourier series of $L^{\infty}(T)$-functions fails to converge in $L^{\infty}$-norm. But does the Fourier series of $f$ converge in $L^{\infty}$-norm for every $f\in C(T)$?

share|cite|improve this question

closed as too localized by Harald Hanche-Olsen, Willie Wong, Andrés Caicedo, Mark Meckes, Yemon Choi Apr 3 '11 at 19:00

This question is unlikely to help any future visitors; it is only relevant to a small geographic area, a specific moment in time, or an extraordinarily narrow situation that is not generally applicable to the worldwide audience of the internet. For help making this question more broadly applicable, visit the help center.If this question can be reworded to fit the rules in the help center, please edit the question.

Partial sums of Fourier series are continuous functions, so you get pointwise convergence everywhere. –  Harald Hanche-Olsen Apr 3 '11 at 12:33
You are confused. $L^\infty$ convergence of continuous functions is uniform convergence. –  Harald Hanche-Olsen Apr 3 '11 at 13:07
No one is disputing that Fourier series can diverge on a null set. Katznelson's example showed that it must be so. The point is that for a continuous function, knowing its behaviour on a set of full measure suffices to prescribe it every where. In other words, $$ \sup_{x\in T} | S_n(s) - S_m(s)| \leq \| S_n - S_m\|_\infty $$ for all partial sums $S_N$. So $L^\infty$ convergence does imply pointwise (and hence even uniform) convergence. What you should think is that for any $a > 0$, the set $A_{n,m}$ of points in $T$ such that $|S_n - S_m| > a$ decays to measure zero as $n,m\to\infty$, but –  Willie Wong Apr 3 '11 at 14:03
Short version: the canonical map from the Banach space $C({\mathbb T})$ into the Banach space $L^\infty({\mathbb T})$ really is injective. –  Yemon Choi Apr 3 '11 at 18:59

2 Answers 2

up vote 3 down vote accepted

If Fourier series of continuous functions would converge in $L^\infty$, then, by the Uniform Boundedness Principle, the operator norms in $C(\mathbb{T})$ of the partial Fourier series operators $S_Nf(t):=\sum_{n=-N}^N\hat{f}(n)e^{int}$ would be uniformly bounded. You can find, for example in Katznelson book, a proof of the fact that such norms diverge logarithmically.

share|cite|improve this answer

I think you can find the answer in P191 "Classical Fourier Analysis",second edition, Loukas Grafakos

share|cite|improve this answer
NO, it's not the answer to the question of mine. The result of duBois Reymond is well known and is a easy corollary of the Banach-Steinhaus theorem. But it is about pointwise convergence, not convergence in $L^\infty$-norm. –  Acky Apr 3 '11 at 13:04
but as the function and the partial sum of the Fourier series are all continuous functions, if we have $L^{\infty}$ convergence of partial sum, then we can get contradiction? –  Shaoming Guo Apr 3 '11 at 13:07

Not the answer you're looking for? Browse other questions tagged or ask your own question.