Basisset not set from freq on an atom at ump2 level


I tried to run the following input file

memory 3000 mb

molecule f {
0 2
f 0.0 0.0 0.0

units angstrom


set basis cc-pVTZ
set reference uhf
set scf_type df
set freeze_core True
set MP2_TYPE df


After received optimization complete, I got output like

Traceback (most recent call last):
File “…psi4”, line 269, in
File “”, line 35, in
File “…”, line 1774, in frequency
H, wfn = hessian(lowername, return_wfn=True, molecule=molecule, **kwargs)
File “…”, line 1448, in hessian
G, wfn = gradient(lowername, molecule=moleculeclone, return_wfn=True, **kwargs)
File “…”, line 638, in gradient
wfn = procedures[‘gradient’][lowername](lowername, molecule=molecule, **kwargs)
File …", line 165, in select_mp2_gradient
return func(name, **kwargs)
File “…”, line 1684, in run_dfocc_gradient
“RIFIT”, core.get_global_option(“BASIS”))
File “…”, line 99, in pybuild_basis
psibasis = core.BasisSet.construct_from_pydict(mol, basisdict, puream)

Fatal Error: CoordEntry::basisset: Basisset not set for F and type of DF_BASIS_CC
Error occurred in file: … on line: 134
The most recent 5 function calls were:

What would be the reason for this issue? notice with ub3lyp method for atom or ri-mp2 for closed-shell molecule, the calculation could run normally. Thanks!

Based on your error message, it looks like you’re using an outdated version of Psi, 1.2 or earlier. The bug you’re reporting strongly reminds me of a bug I fixed in August of last year. Update to the latest version of Psi, 1.3.2, and report back with whether you still get a bug or not. If you do, please provide the updated error message so we can look into it.

Thanks. I updated into 1.3.2. But, received this message for DFT calculation as well

Traceback (most recent call last):
File “/home/quan2019/psi4conda/bin/psi4”, line 287, in
File “”, line 31, in
File “/home/quan2019/psi4conda/lib//python3.6/site-packages/psi4/driver/”, line 1464, in frequency
H, wfn = hessian(name, return_wfn=True, molecule=molecule, **kwargs)
File “/home/quan2019/psi4conda/lib//python3.6/site-packages/psi4/driver/”, line 1308, in hessian
H = driver_findif.compute_hessian_from_gradients(findif_meta_dict, irrep)
File “/home/quan2019/psi4conda/lib//python3.6/site-packages/psi4/driver/”, line 746, in compute_hessian_from_gradients
return _process_hessian(H_pi, B_pi, massweighter, data[“print_lvl”])
File “/home/quan2019/psi4conda/lib//python3.6/site-packages/psi4/driver/”, line 548, in _process_hessian
B = np.vstack(B_blocks)
File “/home/quan2019/psi4conda/lib//python3.6/site-packages/numpy/core/”, line 283, in vstack
return _nx.concatenate([atleast_2d(_m) for _m in tup], 0)

ValueError: need at least one array to concatenate

Printing out the relevant lines from the Psithon --> Python processed input file:
core.set_global_option(“BASIS”, “6-31G*”)
core.set_global_option(“REFERENCE”, “uhf”)
–> frequency(‘b3lyp’)

Psi4 stopped on: Saturday, 29 June 2019 01:01AM
Psi4 wall time for execution: 0:00:05.13

*** Psi4 encountered an error. Buy a developer more coffee!
*** Resources and help at

That error message doesn’t match the input you posted. It says you used the 6-31G* basis, but your input file uses cc-pVTZ. Please post the input you used to produce that error message.

I have a guess as to the problem, but I’ll need to see your input file to confirm. Is the system still only one atom?

My apologize. The input file is

memory 3000 mb

molecule f {
0 2
f 0.0 0.0 0.0

units angstrom


set basis 6-31G*
set reference uhf
#set scf_type df
#set freeze_core True
#set MP2_TYPE df


Alright, thank you.

There are two problems:

  1. The first is a problem on your end. You are asking Psi to compute the vibrational frequencies of a single atom. There are none.
  2. The second is a problem on Psi’s end. Rather than raise an error telling you about (1), it goes along with the computation until it realizes there is no way to proceed. It should have raised an error earlier.

I can fix 2, but the problem here is fundamentally 1. There are no frequencies for a single atom.