Re: [molpro-user] Problem with CC norm

[Kadir Diri]

Hi! I had the same problem in anon calculations and the shift keyword worked for me. I remember I spend a lot of time though, since I tried shift,N where N was 1, 2, 3,  etc; sometimes I had to use 5. And the calculations passed many cycles before they converged successfully with even reduced convergence criteria. In your case it might work better, since my calculations were on dipole-bound anions which are actually sometimes a trouble even for Hartree-Fock . This question has been asked many times, here is a short but a useful explanation: http://www.molpro.net/molpro-user/archive/all/msg00068.html Regards, Kadir Matt Thompson wrote: On Fri, 2004-07-09 at 09:31, Stavros Kardahakis wrote: Dear MOLPRO users, I have some problems with CCSD(T) calculations, both RCCSD(T) and UCCSD(T). In particular, many times I get the error message: "Unreasonable norm. Calculation stopped" and then the job crashes. Is there any way someone can bypass this bottleneck (perhaps ask for another type of normalization) of the CC routine? I am actually getting caught by this too, so I'll double the want for an answer. A search showed a previous question was answered by using uccsd(t);shift,1; but I can't seem to figure out what that does. So, since nothing's worked before, I tried it. Unfortunately, my job has ~3000 sec CC steps (not counting the ~3 h to get to that point), so it'll be a while before I know if this worked. I'm also trying uhf->multi->cc rather than uhf->cc, to see if that makes any difference. Any suggestions from the list? Matt