Timeline for bounds on the entries of an inverse circulant matrix
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 8, 2016 at 19:48 | history | edited | user9072 |
edited tags
|
|
| Dec 4, 2012 at 12:27 | history | edited | Felix Goldberg | CC BY-SA 3.0 |
added 73 characters in body
|
| Dec 4, 2012 at 3:03 | comment | added | fedja | Well, just to see that the matrix is invertible requires checking the values of the polynomial with coefficients given by $d$ at all roots of unity, so that part is unavoidable. After that you have just the standard inverse discrete Fourier transform, so there is hardly anything to simplify in that second part. Do you have some particualr restrictions on $d$ that you think might help? | |
| Dec 2, 2012 at 5:47 | comment | added | Yemon Choi | Not an answer, but a remark: since the algebra of n-by-n circulant matrices is isomorphic to the algebra of polynomials modulo the ideal generated by $X^n-1$, one might try to use norm estimates for polynomials (something $L^2$ flavoured would be my instinctive starting point) | |
| Dec 2, 2012 at 1:38 | history | asked | Felix Goldberg | CC BY-SA 3.0 |