Dear developers,
in some of my test calculations I noticed that CCSD(T) produces stange results (i. e. very far from non-perturbed CC).
When looking at the energy components, I also noticed that "Total CCSD energy" within the CCSD(T) calculation is significantly different from the "Total CCSD energy" from a previous CCSD calculation - even though these values should be clearly identical. I get this strange behavior only sometimes, in most cases the aforementioned two energies are identical, as they should.
In the attached example, what could be the reason for the 0.02 Ha difference in "Total CCSD energies" (between CCSD and CCSD(T))? I checked the input keywords in detail and I could not find anything which could explain such a large deviation.
(I know it is just a rudimentary test with an extremely small basis, but I am still curious)
Thank you in advance,Zsolt
