Dear Xiao, The Hamiltonian and overlap matrices should of course be Hermitian for closed systems. In SIESTA they should be further real symmetric matrices. And indeed they are to within some error.
For a DZ-basis Hamiltonian of 956x956 computed with SIESTA I found: max(abs(H)) = 1.3122, min(abs(H)) = 0 (naturally), and the Frobenius norm of the H-H' matrix (a measure of the Hermiticity) ||H-H'|| = 1.7329e-011 which is pretty small I guess. Also for S the values are max(abs(S)) = 1.0000, min(abs(S)) = 0 and ||S-S'|| = 2.9431e-011. If this is not good enough for you just make them Hermitian by doing a simple H = (H + H')/2 Best regards, Cosmin. :*~*:._.:*~*:._.:*~*:._.:*~*:._.:*~*:._.:*~*: : Cosmin ROMAN, Ph.D. Student. : TIMA-INPG, MNS Group, : 46 Avenue F¨lix Viallet, 38031 Grenoble Cedex, FRANCE. : Tel : 00 33 4 76 57 48 34 : Fax : 00 33 4 76 47 38 14 e-mail : [EMAIL PROTECTED] : :*~*:._.:*~*:._.:*~*:._.:*~*:._.:*~*:._.:*~*: ----- Original Message ----- From: "Xiao Changyong" <[EMAIL PROTECTED]> To: <SIESTA-L@LISTSERV.UAM.ES> Sent: Friday, June 18, 2004 12:19 PM Subject: [SIESTA-L] about the Hamiltonian matrix and orbital overlap matrix > Dear all, > > I met one problem in debugging siesta. I hope your kindly advices. > > I repeated the calculation examples. The system energy can be reproduced as > the same of examples. > However, when I checked the H and S matrices, I found the matrices of H and > S are not symmetry. > In principal, they should be symmetric. Can anyone give me the explanation? > > Thank you very much. > > Best regards, > Xiao, David. > =========================================================== > Changyong XIAO, David > Research Fellow > Department of Chemistry, National Unvieristy of Singapore > 3 Science Drive 3, Singapore 117542 > Tel: (65) 68747880 Fax: (65)67791691 > E-mail: [EMAIL PROTECTED], [EMAIL PROTECTED] > ============================================================