2
$\begingroup$

I am trying to describe the edges of an undirected graph that contains loops. On Wikipedia they are characterized as 2-multisets, meaning it has two elements which can be identical, and the order is not important. I never heard of that term, and find no reference to it anywhere else, so I am not sure whether it is OK to use it in my thesis. Can anyone clear that up for me?

  • 1
    Technically any term that you have well defined can be used in your thesis.2011-01-22
  • 1
    To illustrate what Picakhu said, see http://en.wikipedia.org/wiki/Boojum_%28superfluidity%29 and the Physics Today article by David Mermin linked therein. (Note: using a term and having it taken up by the community seriously are two different things.)2011-01-22
  • 0
    Or, for an older and more widely known term, see http://en.wikipedia.org/wiki/Quark#Etymology2012-06-17
  • 0
    This all seems to assume there can't be more than one edge between two vertices.2012-06-17
  • 0
    @MichaelHardy: Hmm? If there can be multiple edges between a pair of vertices, then the edges of the graph form a multiset rather than a set, but each edge is still an unordered pair (or a "2-multiset", if you prefer).2012-06-17
  • 0
    @IlmariKaronen : I think the set of edges would still be a set, not a multiset, with more than one edge between two vertices. If, for example, one were to assign weights to edges, as is sometimes done, then two edges between a pair of vertices might have different weights. But certainly each edge would still be a "2-multiset".2012-06-17

2 Answers 2

1

Yes. See the Wikipedia article on multisets and the many references therein for where it has occurred previously in literature.

(Note, in the context, a 2-multiset merely means a multiset containing 2 elements.)

1

As Willie Wong notes, it's a perfectly valid term, but at least I personally would find "unordered pair" more natural than "2-multiset".