Exercise 40 - Show in Egoroff's theorem, the hypothesis "$\mu(X)<\infty$" can be replaced by "$|f_n|\le g$ for all $n$, where $g \in L^1(\mu)$."

Egoroff's Theorem - Suppose $\mu(X)<\infty$, and $f_1, f_2, ...$ and $f$ are measurable complex-valued functions on $X$ such that $f_n \rightarrow f$ a.e. Then for every $\epsilon > 0$ there exists $E\subset X$ such that $\mu(E)<\epsilon$ and $f_n \rightarrow f$ uniformly on $E^c$.

The proof of the original Egoroff's theorem uses continuity from above, which requires the finiteness assumption. I tried to use the dominating $g$ to get some finiteness condition.

When $\mu$ is $\sigma$-finite, I showed the integral of $g$ is concentrated on some set of finite measure. But this does not seem to help proving the statement since the complement of this set still have infinite measure (if $X$ has).

Can anyone give me some hint on the problem? Thank you

  • $\begingroup$ The statement that support of $L^1$ function $g$ is of finite measure is false, but instead it is always a $\sigma$-finite set. $\endgroup$ – Song Nov 7 '18 at 2:50

Fix $\varepsilon>0$. For each $N \geq 1$, apply the original version restricted to the set $G_N:=\{g \geq 1/N\}$ (which clearly has finite measure if $\int g \, d\mu<\infty$), in order to get a set $E_N \subset G_N$ with $\mu(E_N)<\frac{\varepsilon}{2^N}$ such that $f_n \to f$ uniformly on $G_N \!\setminus\! E_N$. Now let $E:=\bigcup_{N=1}^\infty E_N$.

Obviously $\mu(E)<\varepsilon$. It's not too difficult to show that $f_n \to f$ uniformly on $X \!\setminus\! E$.

[If you're having trouble with this last bit: First, as an exercise, show that $f_n(x)=\mathbf{1}_{[-n,n]}(x)e^{-|x|}$ converges to $f(x)=e^{-|x|}$ uniformly on $\mathbb{R}$. Then generalise the reasoning.]


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.