Hi, I am using QChem 5.4.0 and trying to do a RASCI calculation with short range version of the VWN functionals. But it keeps giving me the following error:
*Name = SRVWN
Q-Chem fatal error occurred in module libgen/RemStrProc.C, line 134:
I have not used this feature (which is new, I think) but after poking around for a bit, this looks to me like a bug. Frustratingly, there are no sample jobs checked into the code, and none in the manual. I will post a bug ticket about that. Maybe Q-Chem staff could take a look and offer 2nd opinion as to whether I am overlooking something.
Update: setting RAS_SRDFT_COR=SRPBE causes the job to run (i.e., the functional is recognized), but I get a lot of DGEMM errors from BLAS that looks like some other kind of bug. I have posted a ticket (#2633) and cc’d the developer to try to get some clarification.
Thank you. I am trying to reproduce a result from a paper, and according to their method, they used the short range version of VMN. I can have the job run with the VWN5. Because there is a variation between my result and theirs, I don’t know if that was caused by the correlation functionals. But it is strange that this name is not recognized.
It looks to me like a bug or possibly incorrect documentation. Someone will update here if we learn more.
Update: According to the code, the only sr correlation functionals that have recognized names look to be srPBE, srPW92, and srVWN1RPA. For exchange, I see srLSDA and srPBE. This looks like a bug, maybe the complete version of this feature didn’t get checked in. Will update when I know more.
This is right. The available short-range version for the VWN-like correlation is srVWN1RPA. Hence, set RAS_SRDFT_COR to srVWN1RPA.
I’ll provide sample jobs and examples in the manual.