Skip to main content

Timeline for 1 or -1 as an eigenvalue of graph

Current License: CC BY-SA 3.0

10 events
when toggle format what by license comment
Nov 2, 2012 at 8:52 answer added Dima Pasechnik timeline score: 4
Nov 2, 2012 at 8:37 comment added Dima Pasechnik I recall there was a paper by Babai where he related eigenvalues of Cayley graphs with characters of underlying groups: Babai, László Spectra of Cayley graphs. J. Combin. Theory Ser. B 27 (1979), no. 2, 180–189.
Oct 19, 2012 at 0:46 comment added Gordon Royle What is "too large"? There are some pretty good packages out there.
Oct 19, 2012 at 0:37 comment added Aaron Meyerowitz Tell us more about the graph and group. Are there elements of low order? What is their action on the group like?
Oct 18, 2012 at 20:10 comment added Moh514 Many thanks for your answers. Yes, it is cayley graph, and it is too large.
Oct 18, 2012 at 17:55 answer added Aaron Meyerowitz timeline score: 1
Oct 18, 2012 at 11:42 comment added Brendan McKay To continue Chris's comment, if you compute the eigenvector of $\pm 1$ in some small examples, you might see a pattern that you can generalize. Note that an eigenvector permuted according to a group element is also an eigenvector, so probably there will be an eigenvector with not many different entry values. Perhaps one which is constant on the blocks of a block system of the group.
Oct 18, 2012 at 11:17 comment added Chris Godsil The simplest way to show that some number is an eigenvalue is to produce the eigenvector. The second-simplest, if you have the adjacency matrix, is to stuff it into something like sage and compute the factored characteristic polynomial. If you graph is given as a Cayley graph (for example) and is too large for the above strategies, you may be out of luck because then your problem is NP-hard
Oct 18, 2012 at 8:15 answer added Delio Mugnolo timeline score: 2
Oct 18, 2012 at 5:36 history asked Moh514 CC BY-SA 3.0