For example, in the following input
molecule H2 {
0 1
H 0.000000000000 0.000000000000 0.
H 0.000000000000 0.000000000000 0.7414
}
set {
reference rhf
basis cc-pVDZ
freeze_core false
scf_type pk
r_convergence 15
e_convergence 15
d_convergence 15
}
energy('mp2')
I know it is inefficient to do MP2 without density fitting. Nevertheless, I would like to know the result without density fitting anyway. Is that OK to use the CCSD module to produce MP2 as the non-density fitting value?