Comments on What is the reason this example of an everywhere surjective function have or does not have an undefined mean?
Parent
What is the reason this example of an everywhere surjective function have or does not have an undefined mean?
Suppose $f:\mathbb{R}\to\mathbb{R}$ is everywhere surjective (i.e., $f[(a,b)]=\mathbb{R}$ for all non-empty intervals $(a,b)$) where the graph of $f$ has Hausdorff dimension $2$ with zero $2$-d Hausdorff measure—i.e., the measure is defined on the Borel $\sigma$-algebra.
Here is an example.
Question: For each $c,d\in\mathbb{R}$, is the mean of $\mathsf{F}:=\left. f\right|_{(c,d)}$ (Definition $1$) undefined? If so, what is the reason? If not, then what is the reason?
Definition $\S$1. (Mean of $\mathsf{F}$)
Suppose:
- $\dim_{\text{H}}(\cdot)$ is the Hausdorff dimension
- $\mathcal{H}^{\dim_{\text{H}}(\cdot)}(\cdot)$ is the Hausdorff measure in its dimension on the Borel $\sigma$-algebra.
- the integral of $\mathsf{F}$ is defined w.r.t the Hausdorff measure in its dimension
The expected value of $\mathsf{F}:A\to\mathbb{R}$ (i.e., $A:=(c,d)$), w.r.t. the Hausdorff measure in its dimension, is $m_{{}_{\mathsf{F}}\!}(A)$ (when it exists) where:
$$m_{{}_{{\large{\mathsf{F}}}}\!}(A)=\frac{1}{{\mathcal{H}}^{\text{dim}_{\text{H}}(A)}(A)}\int_{A}\mathsf{F}\, d{\mathcal{H}}^{\text{dim}_{\text{H}}(A)}$$
Post
Answer from math stack exchange (I looked up the user online and I am not sure he specializes in mathematics; therefore, take his answer with a pinch of salt):
Yes, the other user is right. The expected value is not undefined because the graph has Hausdorff dimension 2 or zero 2-D Hausdorff measure; that’s irrelevant since expectation integrates over the domain, not the graph. The real reason is that an everywhere-surjective function on every subinterval must be wildly pathological and cannot be Lebesgue-integrable, so the integral defining the expectation fails to exist. The geometry of the graph doesn’t determine integrability; it’s the function’s behavior as a measurable function on (c,d) that matters.
1 comment thread
0 comment threads