F-SAPT error (Fatal Error: Monomer A: A Matrix is not SPD)

SAPT
1

Hello,
I am new to psi4. I am using FSAPT0 (psi4 version 1.3.2) for determining the interactions at transition states. After the CPHF iteration step (the energy and wavefunction of the two monomers were solved excellent). I got the following error.

Traceback (most recent call last):
File “/home/saikat/anaconda/anaconda3/bin/psi4”, line 287, in
exec(content)
File “”, line 75, in
File “/home/saikat/anaconda/anaconda3/lib//python3.7/site-packages/psi4/driver/driver.py”, line 556, in energy
wfn = procedures[‘energy’][lowername](lowername, molecule=molecule, **kwargs)
File “/home/saikat/anaconda/anaconda3/lib//python3.7/site-packages/psi4/driver/procrouting/proc.py”, line 3637, in run_fisapt
fisapt_wfn.compute_energy()
File “/home/saikat/anaconda/anaconda3/lib//python3.7/site-packages/psi4/driver/procrouting/sapt/fisapt_proc.py”, line 71, in fisapt_compute_energy
self.ind()

RuntimeError:
Fatal Error: Monomer A: A Matrix is not SPD
Error occurred in file: /scratch/psilocaluser/conda-builds/psi4-multiout_1557940846948/work/psi4/src/psi4/fisapt/fisapt.cc on line: 4395
The most recent 5 function calls were:

Printing out the relevant lines from the Psithon --> Python processed input file:
core.set_global_option(“MAXITER”, 999)
core.set_global_option(“E_CONVERGENCE”, 1e-8)
core.set_global_option(“D_CONVERGENCE”, 1e-8)
Please could someone explain the error and how I could go about resolving it?

2

You’re not the first one to be confused by that error message.

It means that you landed on the wrong Hartree-Fock state for one of your monomers, that it’s possible to lower the Hartree-Fock energy even more. This is especially common if one of your monomers is a single atom, or when doing open-shell SAPT. I think FSAPT is always closed shell? If so, then the best you can do is set soscf true and add soscf_max_iter 8 starting at 8 but going up to larger numbers if needed. There’s something you can do called “instability following” that would solve the problem better, but we only have that implemented for UHF… We really need to get that for RHF.