I have two functions, and I want to combine them to define a certain function.

Suppose for every fixed $e$ in $(0, ∞)$, we have a function $g_e (x): \mathbb{R} \to [0,\infty]$ that is well defined a.e. and right continuous, and for every fixed x in R, we have a function $h_x (e): (0, ∞) \to [0, ∞]$ that is well defined a.e. and monotone increasing. Further, whenever $e$ and $x$ are such that they are both defined, $g_e (x) = h_x (e)$.

Can we canonically define some $f: \mathbb{R} \times (0, \infty)\to [0,\infty]$ such that it is well defined a.e. and $f(x, e) = g_e (x) = h_x (e)$ a.e.?


At least assuming the continuum hypothesis, the answer is no. Let $\prec$ be a well-order of $\mathbb R$ of order type $\omega_1$. Define $g_e(x)=0$ for $x\succ e$ (and $g_e(x)$ not defined otherwise) and $h_x(e)=e$ if $e\succ x$ (and $h_x(e)$ not defined otherwise). Observe that both $g_e,h_x$ are defined everywhere except at a countable, hence null, set (this is why we need the well-order to have order type $\omega_1$).

Then there is no pair $(x,e)$ for which both $g_e(x),h_x(e)$ are both defined, so the last condition is satisfied vacuously, but there is no function which is equal to both of those numbers simulataneously on any nonempty set.

  • $\begingroup$ Are those g and h necessarily measurable though? The conditions on g and h mean they have to be measurable. $\endgroup$ – James Baxter Dec 25 '18 at 13:52
  • $\begingroup$ @JamesBaxter They are, since they agree with zero, respectively identity, function outside a set of measure zero. $\endgroup$ – Wojowu Dec 25 '18 at 13:53

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.