[molpro-user] MRCC interface

[Rika Kobayashi]

Hello,
Further to my previous mail on this subject (Feb 12 2014), following Kirk's
advice I can confirm that MRCC CC restarts can work quite well by running
the standalone MRCC binaries on the files in the molpro-created directory.
However, we can't get it to work through the molpro input. Ditto with
MRCC's local correlation methods - these look just like a matter of needing
the keyword recognised by molpro?

Is the Molpro interface to MRCC currently being worked on or should we look
into writing some sort of wrapper script? It would be nice to have the MRCC
local coupled cluster method hooked into molpro as the user is currently
evaluating various local coupled cluster methods. Speaking of which, he is
getting from molpro the error:

 in reltripldom: MXDOMT exceeded

where MXDOMT does not appear to be a configurable parameter but fixed to
the number of orbital triples - which is a function of the size of the
calculation (is that right?). Is there a way to get round this error?

Rika

---------------------------------------------------------------------------------------------------------------
Dr Rika Kobayashi

ANU Supercomputer Facility phone: (02) 6125 5986
Australian National University             fax:   (02) 6125 8199
Canberra, ACT 0200,         email: Rika.Kobayashi at anu.edu.au
Australia
http://www.anusf.anu.edu.au/~rxk900
---------------------------------------------------------------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20140324/3f6b54f8/attachment.html>