2
$\begingroup$

Although the following question is not in itself mathematical, it is the expertise/breadth of the research community in mathematics that I wish to appeal to, beyond the filtered/trained search results I'd just get from Googling.


Let $\Omega$ be a non-empty set and ${\mathcal F}$ a non-empty subset of the powerset ${\mathcal P}(\Omega)$. In some work with coauthors we found ourselves frequently needing the following construction: take some $\Gamma\subseteq\Omega$ and form the new set $$ {\mathcal F} \wedge \Gamma := \{ X \cap \Gamma \colon X \in {\mathcal F} \} \subseteq {\mathcal P}(\Gamma). $$ Actually, we then need to take subsets of ${\mathcal F}\wedge \Gamma$; and then iterate this process.

The terminology we are using in the current draft is to call ${\mathcal F}\wedge \Gamma$ the projection of ${\mathcal F}$ onto $\Gamma$, since this seems like the intuitive picture/metaphor. On the other hand, while browsing Bollobas's Combinatorics earlier this year, I noticed that in the setting where $\Omega$ is finite, various people (Frankl, Bollobas himself, presumably others) have called ${\mathcal F}\wedge \Gamma$ the trace of ${\mathcal F}$ on $\Gamma$.

Question. Is the terminology used in Bollobas's book now the standard one? If so, have people seen the notion of "subtrace" before?

For sake of comparison, I don't think anyone would refer to a graph minor as anything but a "graph minor", this is now completely standard as far as I can tell. So my question is really about whether traces and subtraces are just as standard, or whether this is a convenient bit of terminology popular within a particular cluster of researchers. I don't think this is an overly subjective question; I am hoping for combinatorists or lattice theorists to weigh in here with references or informed personal observation.

Another question (less important, more subjective): which is preferable, "subprojection" or "subtrace" ?

$\endgroup$
  • $\begingroup$ For what it's worth, Wikipedia uses "subhypergraph" for your construction: en.wikipedia.org/wiki/Hypergraph#Terminology $\endgroup$ – Sam Hopkins Nov 22 at 17:41
  • 1
    $\begingroup$ @SamHopkins Thanks - this is almost the same although note that my construction allows the empty set to belong to ${\mathcal F}\wedge\Gamma$. Moreover, since in that language I want to delete some vertices and then remove some hyperedges, I seem to have a sub-subhypergraph where the first sub doesn't mean the same as the second sub ... :-/ $\endgroup$ – Yemon Choi Nov 22 at 17:54
  • 1
    $\begingroup$ Right, it does not seem appropriate for your setting, but worth noting. $\endgroup$ – Sam Hopkins Nov 22 at 17:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.