Confusion regarding proof using Fatou's lemma

110 Views Asked by At

This is in reference to the book Problems in Mathematical Analysis III by Kaczor and Nowak. We are given that ${f_n}$ converges to $f$ on $R$. Suppose that $\lim_{n \to \infty} \int_R{f_n dm} - \int_R {f dm} < \infty$. We have to prove that for any measurable A, $\lim_{n \to \infty} \int_A{f_n dm} = \int_A {f dm}$. The proof is using Fatou's lemma and trying to show that $$\int_A {f dm}\leq \liminf_{n \to \infty} \int_A{f_n dm}\leq \limsup_{n \to \infty} \int_A{f_n dm}\leq \limsup_{n \to \infty} \int_R{f_n dm} - \int_{R-A}{f_n dm} \\= \lim_{n \to \infty} \int_{R}{f_n dm} - \liminf_{n \to \infty}\int_{R-A}{f_n dm} .$$

I get it to this point. This is the step that I don't understand. Then say say that $= \lim_{n \to \infty} \int_R{f_n dm} - \liminf_{n \to \infty}\int_{R-A}{f_n dm} $ $\leq \int_R{f dm} - \int_{R-A}{f dm} $

How do they get this? And then this is equal to

$\int_A {f dm}$, which completes the proof. Thanks.

1

There are 1 best solutions below

0
On

It looks like your question boils down to the inequality $$-\liminf_n \int_{\mathbb{R}-A} f_n\, dm \leq -\int_{\mathbb{R}-A} f\, dm$$ But this is nothing more than Fatou's lemma multiplied by $-1$.