TITLE OZONE 6311G* CISD SYMMETRY CNV 2 END VARIABLES ROO 1.272 A THETA 58.4 D END BASIS 6311G* ATOMS OXYGEN 8.0 0.0 0.0 0.0 OXYGEN 8.0 POL ROO THETA 180.0 END CISD ENERGY START FINISHCADPAC also provides a series of CI variants which can be obtained by replacing CISD with the appropriate directive.
results in an averaged coupled-pair functional calculation. ACPF is an approximately size-consistent version of CI.
calculates a coupled electron pair approximation energy, of which there are several variants. The value of the parameter n can be 0,1,2 or 3. The most commonly used forms are CEPA 1 and CEPA 2. CEPA 0 is also known as LCCD (linearised coupled cluster). The CEPA family is another approximately size-consistent variant of CI.
is the simplest of the coupled cluster methods (see next section), including only double excitations.
calculates a quadratic CI (with single and double excitations) energy. QCISD is an exactly size-consistent CISD (or, according to your viewpoint, a simplification of coupled cluster retaining size-consistency).
It should be noted that all the above calculations are approximately
(i.e. within a factor of 2) the same cost. Of this set of methods (CI,
CPF, CEPA, QCI) QCISD is probably the best choice. (For a better method
still, however, see the next section on Brueckner coupled cluster calculations)
which gives a perturbative estimate of the effects of triple excitations. It is noticeably better than QCISD, but also more expensive.
The output, including the iterative part of the calculation, will look like,
RUNNING A QUADRATIC CI CALCULATION ITERATION 1 CORRELATION ENERGY -0.6265003533 TOTAL ENERGY -224.8752875086 ITERATION 2 CORRELATION ENERGY -0.5597977701 TOTAL ENERGY -224.8085849255 . . ITERATION 13 CORRELATION ENERGY -0.6052211693 TOTAL ENERGY -224.8540083247 ITERATION 14 CORRELATION ENERGY -0.6052212571 TOTAL ENERGY -224.8540084125This is the result of a QCISD calculation for the ozone example. If the triples had been included as well i.e. QCISD(T), then the following extra output would be produced.
EVALUATING THE TRIPLES CONTRIBUTION Running vectorised medium memory version Triples contribution -0.0304561390 TOTAL Energy -224.8844645515
where n is an integer. This results in the convergence on the energy
being set to 10-n. The default value is n=6. This is
adequate for energies. However, if one is taking differences of energies,
for example, when working out polarisabilities by finite difference (see
section 6.x) then it would be advisable to converge the calculation more
strictly than the default.