I’m using the Psi4 Linux executables downloaded from your site. It seems that passing a conventional MP2 wave function to the fchk() facility causes a core dump. The following is a simplified example input to trigger the problem:
molecule test1 {
0 1
Ne
}
set basis cc-pvdz
set mp2_type conv
energy, wfn = energy(‘mp2’, return_wfn=True)
fchk_writer = psi4.FCHKWriter(wfn)
fchk_writer.write(‘test1.fchk’)
This works fine for mp2_type = cd or df.
Is there a workaround for this? I just need to do a test of the conventional mp2 wave function (for a larger molecule; I don’t actually need the mp2 density). What about writing out the wfn to a file in some format that can be then read in as an initial guess, doing an SCF with 0 iterations, and then using fchk()? Is there a way to do that?
regards,
Dale