Incorrect Implementation of HamPPP for 1-electron integrals

[mdtoo]

Hello,

I discovered an error in the class HamPPP in the file hamiltonians.py when comparing against my own PPP code that runs SCF calculations on a lattice with different values for the nuclear charge. The off-site coulomb interaction term in the PPP Hamiltonian distributes as

for which the 2nd term shows that the diagonal element p should be associated with the charge q in the 1-electron integrals. The code snippet below uses the charge p instead, which is an incorrect implementation.

Switching p and q provides correct agreement with my own codebase.