2
$\begingroup$

Let $f$ be strictly increasing on $\mathbb{R}$. Then $x \oplus y := f^{-1}(f(x)+f(y))$ gives rise to a strict symmetric monoidal ($\Rightarrow$ commutative monoid) structure on $(\mathbb{R},\ge)$ with monoidal unit $f^{-1}(0)$. If $f(x) := \exp(x/h)$ for $h > 0$, then $([-\infty,\infty],\oplus,-\infty,+,0)$ is "the" widely used log semiring.

If $f(x) := \sigma(x)|x|^p$ for $p > 0$, then $(\mathbb{R},\oplus,0,\cdot,1)$ is also a semiring.

This semiring does not appear to be widely used, and it is not well-known to me. What is it called, and where can I find references to it in the literature?

NB. I imagine that both of these semirings are morally unified under the aegis of "associated homogeneity" in the sense of Gel'fand and Shilov, i.e., with $\log$ a variant of homogeneous of degree zero.

$\endgroup$
5
  • 1
    $\begingroup$ What is $\sigma(x)$? $\endgroup$ Commented Feb 18, 2021 at 16:57
  • 1
    $\begingroup$ @AbdelmalekAbdesselam It's en.wikipedia.org/wiki/Sign_function $\endgroup$ Commented Feb 18, 2021 at 17:32
  • 3
    $\begingroup$ It would be better to edit a definition of $\sigma(x)$ into the body of the question, than to make users chase it down offsite. $\endgroup$ Commented Mar 10, 2023 at 21:36
  • 2
    $\begingroup$ I'm confused about this question: it seems to me that, yes, this semiring, which is in fact a ring, is very widely used since it is simply (isomorphic to) the ring $\mathbb{R}$ with its usual addition and multiplication, the bijection $f$ being the isomorphism. Obviously for any mathematical structure $X$ and bijection $f$ from a different set $Y$ we can transfer the structure on $X$ to $Y$, but it's still the same structure! Now maybe there's still some reason to distinguish two isomorphic structures, but if so, it should have been clarified in the question. $\endgroup$
    – Gro-Tsen
    Commented Mar 10, 2023 at 21:51
  • $\begingroup$ @Gro-Tsen Along with the exponential, this particular choice of $f$ is relevant for certain applications and actual numerical computations will be nice. A hint of this is on page 2 of arxiv.org/abs/2205.05178. If you have seen this particular choice of $f$ used I would love to hear about it. $\endgroup$ Commented Mar 11, 2023 at 16:55

1 Answer 1

1
$\begingroup$

This is not an answer to the OP, but is too long for a comment.

You may want to check Example (1.23) in J.S. Golan's Semirings and their Applications (Springer, 1999). The first sentence reads as follows:

If $(R, +, \cdot\,)$ is a semiring and $X$ is a set together with a bijective function $\delta \colon X \to R$, then the semiring structure on $R$ induces a semiring structure $(X, \oplus, \odot)$ on $X$ with the operations defined by $x \oplus y = \delta^{-1}(\delta(x) + \delta(y))$ and $x \odot y = \delta^{-1}(\delta(x) \cdot \delta(y))$.

Golan mentions that a special case of this construction is considered by A.A. Mullin in: On the algebraic structure of pre-rings, Notices Amer. Math. Soc. 22 (1975), A703.

$\endgroup$
3
  • $\begingroup$ scholar.google.com/… shows nothing at all. $\endgroup$ Commented Feb 18, 2021 at 17:39
  • $\begingroup$ I haven't seen Mullin's paper, but it's almost surely an "extended abstract". The 1975 volume of the Notices is freely available: ams.org/cgi-bin/notices/… (unfortunately, Golan only gives the volume number). $\endgroup$ Commented Feb 18, 2021 at 17:51
  • $\begingroup$ It's on p. 47 of the November issue, and about the continued fraction stuff that Golan mentions. $\endgroup$ Commented Feb 18, 2021 at 18:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .