Karl
September 7, 2016, 7:49am
1
Hello again,
I have ran into an error concerning the CBS keyword. Using the following input:
cbs(‘mp2’, corl_basis=‘aug-cc-pv[q5]z’, corl_scheme=corl_xtpl_helgaker_2)
I receive the following error message:
“TypeError: complete_basis_set() takes exactly 2 arguments (0 given)”.
This error message is also given if I use any of the cbs examples that are provided in the manual. Note that the calculation works if I use an older version of Psi4 (i.e. Psi4 0.4.186)
All the bests,
loriab
September 7, 2016, 5:16pm
2
Sorry about that. The docs need some redirecting for the user. The cbs(...) function is powerful but complicated because of all those options. We rearranged so that users interact directly with fns like energy(), optimize(), frequency() and only indirectly though cbs(), cp(), etc.
So, your calc can be accessed through
energy('mp2/aug-cc-pv[q5]z')
Or, if you really like the arguments,
energy(cbs, corl_wfn='mp2', corl_basis='aug-cc-pv[q5]z', corl_scheme=corl_xtpl_helgaker_2).
Some more examples are at the files below.
#! Extrapolated energies with delta correction
nucenergy_ref = 5.1767335826 #TEST
molecule hf {
F
H 1 0.92
}
# Get a reasonable guess, to save some iterations
set {
scf_type df
mp2_type df
e_convergence 7
reference rhf
}
hf.update_geometry()
compare_values(nucenergy_ref, hf.nuclear_repulsion_energy(), 9, "Nuclear repulsion energy") #TEST
show original
#! optimization with method defined via cbs
# NOTE: gradient calcs accessed through an alias to cbs (e.g., sherrill_gold_standard) aren't matching direct calls to cbs
molecule mol {
O
H 1 R
H 1 R 2 A
}
mol.R = 1.0
mol.A = 90.0
set g_convergence gau_verytight
E1 = energy(cbs, corl_wfn='mp2', corl_basis='cc-pV[DT]Z', delta_wfn='ccsd(t)', delta_basis='3-21g')
E2 = energy(sherrill_gold_standard, scf_basis='cc-pVTZ', corl_basis='cc-pV[DT]Z', delta_basis='3-21g')
compare_values(E1, E2, 5, '[1] Match gold_standard energy') #TEST
G1 = gradient(cbs, corl_wfn='mp2', corl_basis='cc-pV[DT]Z', delta_wfn='ccsd(t)', delta_basis='3-21g', dertype='energy')
show original
Karl
September 8, 2016, 11:11am
3
Hello Lori - great and thanks for the info. The calculations all works now.
Bests,