1

Notation)

  • $\mathbf{x} = (x_1, \cdots, x_n)$, $\mathbf{y} = (y_1, \cdots, y_m)$,
  • $I_1=\{\mathbf{x}: a_i\le x_i\le b_i, ~~i=1, \cdots, n\}$
  • $I_2=\{\mathbf{y}: c_j\le y_j\le d_j, ~~j=1, \cdots, m\}$,
  • $\displaystyle L(E)=\left\{f:\int_E f \quad\text{is finite}\right\}$
  • $L(d\mathbf{x})=L(I_1)$, $L(d\mathbf{y})=L(I_2)$

Theorem $6.1$ (Fubini's Theorem)

Let $f(\mathbf{x}, \mathbf{y})\in L(I), I=I_1\times I_2.$ Then

  • For almost every $\mathbf{x}\in I_1$, $f(\mathbf{x}, \mathbf{y})$ is measurable and integrable on $I_2$ as a function of $\mathbf{y}$
  • As a function of $\mathbf{x}$, $\int_{I_2} f(\mathbf{x}, \mathbf{y}) d\mathbf{y}$ is measurable and integrable on $I_1$, and $$ \iint_I f(\mathbf{x}, \mathbf{y}) d\mathbf{x}d\mathbf{y} = \int_{I_1} \left[ \int_{I_2} f(\mathbf{x}, \mathbf{y}) d\mathbf{y} \right] d\mathbf{x} $$

My textbook says: In these lemmas, we say that a function $f$ in $L(d\mathbf{x}d\mathbf{y})$ for which Fubini's theorem is true has property $\mathfrak{F}$.


Lemma $6.2$

A finite linear combination of functions with property $\mathfrak{F}$ has property $\mathfrak{F}$.

proof) This follows immediately from Theorems $4.9$ and $5.28$.


Theorem $4.9$ $$f \text{ and } g \text{ are measurable.}\Longrightarrow f+g \text{ is measurable.}$$

Theorem $5.28$ $$f,g\in L(E) \Longrightarrow \cases{f+g\in L(E) \\ \\ \displaystyle\int_E (f+g) = \int_E f + \int_E g}$$


Now, my question is

$1$. What does property $\mathfrak{F}$ exactly mean? I failed to analyze the following sentense: "we say that a function $f$ in $L(d\mathbf{x}d\mathbf{y})$ for which Fubini's theorem is true has property $\mathfrak{F}$." Is $f$ said to have property $\mathfrak{F}$ if $f\in L(d\mathbf{x}d\mathbf{y})$?

$2$. If so, in order to prove lemma $6.2$, theorem $4.9$ does not seem to be needed. Only theorem $5.28$ is enough to prove lemma $6.2$, isn't it?


p.s. I don't know which tags I should take. Heretofore, I tagged real analysis, lebesgue integral, lebesgue measure, measure theory, and so on. However, people often edited my tags. So, I just tagged Lebesgue Integral in this post.

Danny_Kim
  • 3,423

1 Answers1

2

I may be misreading what you've written, but I believe that the property is simply: If the results of Fubini's Theorem holds for f, then f is said to have that property.

I believe the author is just using it for notational convenience. Which text are you working with, that might help us confirm this.

David
  • 3,180
  • Thanks for answering my question. I am working with "Measure and Integral: An introduction to real analysis" by Wheeden and Zygmund. – Danny_Kim May 27 '16 at 16:13
  • If $\mathfrak{F}$ means what you said, why does the author check measurability and integrability? Furthermore, I think all integrable function is already measurable although converse does not hold, isn't it? – Danny_Kim May 27 '16 at 16:15
  • 1
    Because the results of fubini's theorem holding for a function does not imply that it met the sufficient conditions for it to hold. The conditions given in Thm 6.1 are not necessary and sufficient, just sufficient.

    I have confirmed that what I stated originally is correct, see Section 8 in: http://www.math.rochester.edu/people/grads/clungstr/real_analysis/Notes.pdf

    for example.

    – David May 27 '16 at 16:21
  • The link is very useful and helpful. Thank you. However, it does not resolve the curiosity, which is why the author checks measurability and integrability. Only checking integrability is enough since if $f$ is integrable, $f$ is automatically measurable, isn't it? – Danny_Kim May 27 '16 at 16:38
  • 1
    Sorry, I'm getting on a plane so I don't have time to look too closely, but I believe that thm 4.9 is necessary to demonstrate that it holds for linear combinations which are measurable, but not integrable. 5.28 is sufficient if we are only concerned with integrable functions. – David May 27 '16 at 16:54