=================== Perturbation Module =================== All subroutines in `*.irp.f` starting with ``pt2_`` in the current directory are perturbation computed using the routine ``i_H_psi``. Other cases are not allowed. The arguments of the ``pt2_`` are always: subroutine pt2_...( & psi_ref, & psi_ref_coefs, & E_refs, & det_pert, & c_pert, & e_2_pert, & H_pert_diag, & Nint, & ndet, & n_st ) integer, intent(in) :: Nint,ndet,n_st integer(bit_kind), intent(in) :: psi_ref(Nint,2,ndet) double precision , intent(in) :: psi_ref_coefs(ndet,n_st) double precision , intent(in) :: E_refs(n_st) integer(bit_kind), intent(in) :: det_pert(Nint,2) double precision , intent(out) :: c_pert(n_st),e_2_pert(n_st),H_pert_diag psi_ref bitstring of the determinants present in the various n_st states psi_ref_coefs coefficients of the determinants on the various n_st states E_refs Energy of the various n_st states det_pert Perturber determinant c_pert Pertrubative coefficients for the various states e_2_pert Perturbative energetic contribution for the various states H_pert_diag Diagonal H matrix element of the perturber Nint Should be equal to N_int Ndet Number of determinants `i` in Psi on which we apply N_st Number of states Assumptions =========== .. Do not edit this section. It was auto-generated from the .. NEEDED_MODULES file. * This is not allowed: subroutine & pt2_.... Documentation ============= .. Do not edit this section. It was auto-generated from the .. NEEDED_MODULES file. `pt2_moller_plesset `_ compute the standard Moller-Plesset perturbative first order coefficient and second order energetic contribution .br for the various n_st states. .br c_pert(i) = /(difference of orbital energies) .br e_2_pert(i) = ^2/(difference of orbital energies) .br `pt2_epstein_nesbet `_ compute the standard Epstein-Nesbet perturbative first order coefficient and second order energetic contribution .br for the various n_st states. .br c_pert(i) = /( E(i) - ) .br e_2_pert(i) = ^2/( E(i) - ) .br `pt2_epstein_nesbet_2x2 `_ compute the Epstein-Nesbet 2x2 diagonalization coefficient and energetic contribution .br for the various n_st states. .br e_2_pert(i) = 0.5 * (( - E(i) ) - sqrt( ( - E(i)) ^2 + 4 ^2 ) .br c_pert(i) = e_2_pert(i)/ .br `pt2_epstein_nesbet_2x2_sc2 `_ compute the Epstein-Nesbet 2x2 diagonalization coefficient and energetic contribution .br for the various n_st states. .br but with the correction in the denominator .br comming from the interaction of that determinant with all the others determinants .br that can be repeated by repeating all the double excitations .br : you repeat all the correlation energy already taken into account in CI_electronic_energy(1) .br that could be repeated to this determinant. .br ---> + delta_e_corr .br e_2_pert(i) = 0.5 * (( - E(i) ) - sqrt( ( - E(i)) ^2 + 4 ^2 ) .br c_pert(i) = e_2_pert(i)/ .br `pt2_epstein_nesbet_sc2 `_ compute the Epstein-Nesbet perturbative first order coefficient and second order energetic contribution .br for the various n_st states, .br but with the correction in the denominator .br comming from the interaction of that determinant with all the others determinants .br that can be repeated by repeating all the double excitations .br : you repeat all the correlation energy already taken into account in CI_electronic_energy(1) .br that could be repeated to this determinant. .br ---> + delta_e_corr .br c_pert(i) = /( E(i) - ( ) ) .br e_2_pert(i) = ^2/( E(i) - ( ) ) .br `pt2_epstein_nesbet_sc2_projected `_ compute the Epstein-Nesbet perturbative first order coefficient and second order energetic contribution .br for the various n_st states, .br but with the correction in the denominator .br comming from the interaction of that determinant with all the others determinants .br that can be repeated by repeating all the double excitations .br : you repeat all the correlation energy already taken into account in CI_electronic_energy(1) .br that could be repeated to this determinant. .br BUT on the contrary with ""pt2_epstein_nesbet_SC2"", you compute the energy by projection .br ---> + delta_e_corr .br c_pert(1) = 1/c_HF /( E(i) - ( ) ) .br e_2_pert(1) = c_pert(1) .br NOTE :::: if you satisfy Brillouin Theorem, the singles don't contribute !! .br `cisd `_ Undocumented `fill_h_apply_buffer_selection `_ Fill the H_apply buffer with determiants for the selection `selection_criterion `_ Threshold to select determinants. Set by selection routines. `selection_criterion_factor `_ Threshold to select determinants. Set by selection routines. `selection_criterion_min `_ Threshold to select determinants. Set by selection routines. `n_det_ref `_ Undocumented `psi_ref `_ On what we apply for perturbation. If selection, it may be 0.9 of the norm. `psi_ref_coef `_ On what we apply for perturbation. If selection, it may be 0.9 of the norm. `psi_ref_size `_ Undocumented Needed Modules ============== .. Do not edit this section. It was auto-generated from the .. NEEDED_MODULES file. * `AOs `_ * `BiInts `_ * `Bitmask `_ * `Dets `_ * `Electrons `_ * `Ezfio_files `_ * `Hartree_Fock `_ * `MonoInts `_ * `MOs `_ * `Nuclei `_ * `Output `_ * `Utils `_ * `SingleRefMethod `_