Hello all developers of Pyscf
I have calculated the MO-coeff matrix by kHF of diamond, some elements appear different at each time of calculation, the difference can be higher than 10^-4. The difference of ERI in kmp2 can be observed but the energy of kmp2 or kHF still get the same value. Can someone please tell me why the difference exist and why the energy of kmp2 and kHF unchanged. Thank you!!!

This is my input file

import pyscf.pbc.tools.pyscf_ase as pyscf_ase
import pyscf.pbc.gto as pbcgto
from pyscf.pbc import gto, scf, cc, mp
from ase.lattice import bulk

c = bulk('C', 'diamond', a=3.5668)
print(c.get_volume())

cell = pbcgto.Cell()
cell.atom = pyscf_ase.ase_atoms_to_pyscf(c)
cell.a = c.cell

cell.basis = 'gth-dzv'
cell.pseudo = 'gth-pade'
cell.verbose = 5
cell.build(None,None)

kpts = cell.make_kpts([1,2,1], scaled_center=[0., 0., 0.])
kmf = scf.KRHF(cell, kpts)
kmf.kernel()
print(kmf.mo_coeff)

mypt = mp.KMP2(kmf)
mypt.kernel()