Show that the pointwise limit of integrable functions is not necessarily integrable.

I am stuck on this question. Here is what I know.

Let $(f_n)^{\infty}_{n=1}$ be a series of integrable functions, and let

$$\lim_{n \to \infty} f_n(x)=Z$$

I need to show that $Z$ is not necessarily integrable. Should I be looking for a specific example? A function that is integrable, but as $n \to \infty $ the function is no longer integrable.

  • $\begingroup$ You should be a bit more specific by what you mean by integrable. $\endgroup$ – xavierm02 Mar 15 '13 at 12:35
  • $\begingroup$ To answer your question, yes. Check out dominated convergence theorem for details. $\endgroup$ – Gautam Shenoy Mar 15 '13 at 13:17

Here is another natural example on $\mathbb{R}$ with the Lebesgue measure, which works also on $\mathbb{Z}$ with the counting measure: $$ f_n=1_{[-n,n]} $$ Use the Monotone Convergence Theorem.

In the same spirit, we can do even worse. Take $$ g_n=-1_{[-n,0)}+1_{(0,n]}. $$ Then all the $g_n$'s have $0$ integral. So $\lim _n \int g_n=0$ exists. Yet the pointwise limit is not integrable. Again neither on $\mathbb{R}$ nor on $\mathbb{Z}$.


Just in case your question is about Riemann integrability, I will provide a sequence of Riemann integrable functions $f_n\colon[0,1]\to\mathbb{R}$ such that $\lim_{n\to\infty}f_n(x)=f(x)$ exists for all $x\in[0,1]$ but $f$ is not integrable in the sense of Riemann.

Order the rationals in $[0,1]$: $$ [0,1]\cap\mathbb{Q}=\{r_n\}_{n=1}^\infty. $$ Let $$ \phi_n(x)=\begin{cases} 1 &\text{if }x=r_n,\\ 0 &\text{otherwise} \end{cases} \quad\text{and}\quad f_n(x)=\sum_{k=1}^n\phi_k(x). $$ Then $f_n$ is Riemann integrable, since it has a finite number of discontinuities ($r_1,\dots,r_n$), but $$ f(x)=\lim_{n\to\infty}f_n(x)=\sum_{k=1}^\infty\phi_k(x)=\begin{cases} 1 &\text{if }x\text{ is rational}\\ 0 &\text{if }x\text{ is irrational} \end{cases} $$ is not Riemann integrable (althoug it is Lebesgue integrable).

  • $\begingroup$ Aguirre: Might be a stupid question. Can you please explain the last equality. Thanks $\endgroup$ – Mr. MBB Jul 29 '16 at 1:50
  • $\begingroup$ If $x$ is irrational then $\phi_k(x)=0$ for all $k$ and the sum is $0$. If $x$ is rational then $x=r_n$ for some $n$, and all the terms in the sum are $0$ except the corresponding to $k=n$ which is equal to $1$. $\endgroup$ – Julián Aguirre Jul 29 '16 at 8:36

Yes, a specific example is what you need, and you might start by looking at a well-known function which is not integrable (for example $f(x)=1/x$) and then find some integrable functions which converge to $f$ - e.g. consider the functions $f_n(x) = \max(f(x),n)$. I will let you check the details ...


Check this,

$$ \int_{0}^{1}x^{-1+\frac{1}{n}}dx. $$

Note that, the functions $x^{1+1/n}$ are integrable as improper integrals or in the extended sense of Riemann and Lebesgue.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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