[Edited thanks to Daniel's comment.]
The following is the Hahn decomposition theorem stated in Folland's Real Analysis:
The key step, which could be called as the "greedy algorithm", is as follows
I would like see how this theorem works for a particular example. Suppose $$ \nu(E):=\int_Ef\ d\mu $$ where $\mu$ is a positive measure on the measurable space $(X,\mathcal{M})$ and $f$ is an extended $\mu$-integrable function.
[Added: one should note that $-\infty$ in the second line of the proof is a typo and it should be $+\infty$ instead.]
Here is my question:
How can one describe the Hahn decomposition for $\nu$ in terms of $f$ and $\mu$? (This is also Exercise 6 in Chapter 3 of Folland's Real Analysis.)

