27 Sep
2008
27 Sep
'08
10:21 p.m.
On 9/27/08, victor miller <victorsmiller@gmail.com> wrote:
... Define
F[j] := sum_j S[i]*w^{i j}
Then the determinant is prod_j F[j].
Should I think have read F[j] := sum_k S[k]*w^{k j} Good point about case j = 0. In fact, I did start off looking at {0,1} circulants, which displayed similar unexpected behaviour --- though to a lesser extent, for the reason mentioned. Edwin Clark has also communicated to me a proof for the prime case, and placed a list of counts for low n at OEIS A144926. WFL