I need a hand with the following exercise:
Prove that $f: (0,2) \to \mathbb{R}$ given by
$f(x) = \begin{cases} \frac{1}{x} & 0<x\leq 1 \\ \frac{1}{x-2} & 1< x < 2 \end{cases} $
Is not Lebesgue-integrable.
Here are my thought:
We can write $f$ as $f = f^+ - f^-$ where
$f^+(x) = \begin{cases} f(x) & \text{if} \ f(x)>0 \\ 0 &\text{otherwise}& \end{cases} $
$f^-(x) = \begin{cases} -f(x) & \text{if} \ f(x)\le0 \\ 0 &\text{otherwise}& \end{cases} $
And by definition $f$ is integrable if and only if $f^+$ and $f^-$ are both integrable, so I just need to prove that $f^+$ or $f^-$ is not Lebesgue integrable.
Now, in this particular case $f^+(x) = \begin{cases} \frac{1}{x} & 0<x\leq1 \\ 0 & 1<x<2 \\ \end{cases} $ and $f^-(x) = \begin{cases} 0 & 0<x<1 \\ -\frac{1}{x-2} & 1\leq x<2 \\ \end{cases} $
So $f = f^+-f^-$ becomes $$ f = \frac{1}{x}\chi_{(0,1]}-\left(-\frac{1}{x-2}\chi_{[1,2)}\right)$$
thus if I proove that, says, $\frac{1}{x}$ is not Lebesgue integrable on $(0,1]$ the problem is solved. But how to prove that?
$\int |f| \geq \int_{1/n}^{1} \frac 1 x\, dx=\log \, n \to \infty$ so $f$ is not Lebesgue integrable. I have used the fact that on $(\frac 1n, 1)$ the function is Riemann integrable which implies that it is Lebesgue intergable and the two integrals are equql.