Positive function with finite integral is finite almost surely

499 Views Asked by At

Let $\displaystyle f:(X,\Sigma, \mu)\to (\bar{\mathbb R_+}, \mathcal B)$ be a measurable non-negative function that may assume the value $+\infty$.

Suppose that $\int f d\mu$ is finite. Prove that $\{x\in X, f(x)=+\infty\}$ has measure $0$.

Here's my proof.

Let $A=\{x\in X, f(x)=+\infty\}$.

Suppose for contradiction that $\mu(A)>0$ and let $M>0$.

Then $f\geq 1_{A}M$. Hence $\int f d\mu\geq M \mu(A)$

Letting $M$ go to $\infty$ yields a contradiction.

Since I'm new to measure-theory, I'd like to know if there are shorter, different or more conventional proofs.

1

There are 1 best solutions below

1
On BEST ANSWER

A different proof: By Markov's inequality,

$$\mu(\{f \geq R\}) \leq \frac{\int f \, d\mu}{R}.$$

Hence, by the continuity of the measure $\mu$,

$$\mu(\{f=\infty\}) = \lim_{R \to \infty} \mu(\{f \geq R\}) = 0.$$


Another one: From Tonelli's Theorem it is not difficult to see that

$$\int f \, d \mu = \int_{(0,\infty)} \mu(\{f \geq R\}) \, dR$$

for any non-negative (measurable) function $f$. Since the left-hand side is finite and $R \mapsto \mu(\{f \geq R\})$ decreasing, this implies

$$\lim_{R \to \infty} \mu(\{f \geq R\})=0;$$

hence, $\mu(\{f=\infty\})=0$.