1
$\begingroup$

I am interested in convergence results that are true for Riemann-integrable functions but fail for Lebesgue-integrable functions. I know three of these, which happen to be closely related.

  • Convergence of Riemann sums of a Riemann-integrable function $f : [a,b] \rightarrow {\bf R}$.

  • Convergence of means of Riemann-integrable functions along equidistributed sequences (Weyl criterion).

  • The weak convergence of $P_n$ to $P$ implies $\int f dP_n \rightarrow \int f dP$ for all $P$-Riemann-integrable $f$ ($P$-Riemann integrability meaning that the set of discontinuities of $f$ is negligible wrt $P$).

Any others?

$\endgroup$
0

1 Answer 1

2
$\begingroup$

As I learned and discussed here, if $f\colon\mathbb{R}\to\mathbb{R}$ is Riemann-integrable, and if $\mathscr{M}_n(f) := \frac{1}{n}\sum_{k=0}^{n-1}\; f\big(\!x+\frac{k}{n}\big)$ denotes the average of the translates of $f$ by multiples of $1/n$, then $\mathscr{M}_n(f)$ converges uniformly to the average value $\int_0^1 f$ of $f$ (a constant function), but convergence a.e. does not hold for $f$ Lebesgue-integrable, even if it is assumed to be bounded (Walter Rudin, "An Arithmetic Property of Riemann Sums", Proc. Amer. Math. Soc. 15 (1964), 321–324).

$\endgroup$
2
  • $\begingroup$ I would say this belongs to the item "convergence of Riemann sums". Thanks for the reference that looks interesting. $\endgroup$
    – coudy
    Jun 29, 2016 at 19:45
  • 2
    $\begingroup$ @coudy It is of course a kind of convergence of Riemann sums, but note that for $f$ R-integrable we get uniform convergence, so it's a bit stronger in that sense; and, of course, the negative result for $f$ L-integrable is also not subsumed by general nonconvergence because this is a very particular kind of Riemann sums. $\endgroup$
    – Gro-Tsen
    Jun 29, 2016 at 20:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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