10
0
mirror of https://github.com/QuantumPackage/qp2.git synced 2024-11-15 10:33:50 +01:00
QuantumPackage/plugins/local/non_h_ints_mu/new_grad_tc.irp.f

172 lines
5.5 KiB
Fortran

! ---
BEGIN_PROVIDER [double precision, tc_grad_and_lapl_ao_loop, (ao_num, ao_num, ao_num, ao_num)]
BEGIN_DOC
!
! tc_grad_and_lapl_ao_loop(k,i,l,j) = < k l | -1/2 \Delta_1 u(r1,r2) - \grad_1 u(r1,r2) . \grad_1 | ij >
!
! = 1/2 \int dr1 (phi_k(r1) \grad_r1 phi_i(r1) - phi_i(r1) \grad_r1 phi_k(r1)) . \int dr2 \grad_r1 u(r1,r2) \phi_l(r2) \phi_j(r2)
!
! This is obtained by integration by parts.
!
END_DOC
implicit none
integer :: ipoint, i, j, k, l
double precision :: weight1, contrib_x, contrib_y, contrib_z, tmp_x, tmp_y, tmp_z
double precision :: ao_k_r, ao_i_r, ao_i_dx, ao_i_dy, ao_i_dz
double precision :: ao_j_r, ao_l_r, ao_l_dx, ao_l_dy, ao_l_dz
double precision :: time0, time1
double precision, allocatable :: ac_mat(:,:,:,:)
print*, ' providing tc_grad_and_lapl_ao_loop ...'
call wall_time(time0)
allocate(ac_mat(ao_num,ao_num,ao_num,ao_num))
ac_mat = 0.d0
! ---
do ipoint = 1, n_points_final_grid
weight1 = 0.5d0 * final_weight_at_r_vector(ipoint)
do i = 1, ao_num
ao_i_r = weight1 * aos_in_r_array (i,ipoint)
ao_i_dx = weight1 * aos_grad_in_r_array(i,ipoint,1)
ao_i_dy = weight1 * aos_grad_in_r_array(i,ipoint,2)
ao_i_dz = weight1 * aos_grad_in_r_array(i,ipoint,3)
do k = 1, ao_num
ao_k_r = aos_in_r_array(k,ipoint)
tmp_x = ao_k_r * ao_i_dx - ao_i_r * aos_grad_in_r_array(k,ipoint,1)
tmp_y = ao_k_r * ao_i_dy - ao_i_r * aos_grad_in_r_array(k,ipoint,2)
tmp_z = ao_k_r * ao_i_dz - ao_i_r * aos_grad_in_r_array(k,ipoint,3)
do j = 1, ao_num
do l = 1, ao_num
contrib_x = int2_grad1_u12_ao(l,j,ipoint,1) * tmp_x
contrib_y = int2_grad1_u12_ao(l,j,ipoint,2) * tmp_y
contrib_z = int2_grad1_u12_ao(l,j,ipoint,3) * tmp_z
ac_mat(k,i,l,j) += contrib_x + contrib_y + contrib_z
enddo
enddo
enddo
enddo
enddo
! ---
do j = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do k = 1, ao_num
tc_grad_and_lapl_ao_loop(k,i,l,j) = ac_mat(k,i,l,j) + ac_mat(l,j,k,i)
enddo
enddo
enddo
enddo
deallocate(ac_mat)
call wall_time(time1)
print*, ' Wall time for tc_grad_and_lapl_ao_loop = ', time1 - time0
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, tc_grad_and_lapl_ao, (ao_num, ao_num, ao_num, ao_num)]
BEGIN_DOC
!
! tc_grad_and_lapl_ao(k,i,l,j) = < k l | -1/2 \Delta_1 u(r1,r2) - \grad_1 u(r1,r2) . \grad_1 | ij >
!
! = -1/2 \int dr1 (phi_k(r1) \grad_r1 phi_i(r1) - phi_i(r1) \grad_r1 phi_k(r1)) . \int dr2 \grad_r1 u(r1,r2) \phi_l(r2) \phi_j(r2)
! = 1/2 \int dr1 (phi_k(r1) \grad_r1 phi_i(r1) - phi_i(r1) \grad_r1 phi_k(r1)) . \int dr2 (-1) \grad_r1 u(r1,r2) \phi_l(r2) \phi_j(r2)
!
! -1 in \int dr2
!
! This is obtained by integration by parts.
!
END_DOC
implicit none
integer :: ipoint, i, j, k, l, m
double precision :: weight1, ao_k_r, ao_i_r
double precision :: time0, time1
double precision, allocatable :: b_mat(:,:,:,:)
print*, ' providing tc_grad_and_lapl_ao ...'
call wall_time(time0)
if(read_tc_integ) then
open(unit=11, form="unformatted", file=trim(ezfio_filename)//'/work/tc_grad_and_lapl_ao', action="read")
read(11) tc_grad_and_lapl_ao
close(11)
else
PROVIDE int2_grad1_u12_ao
allocate(b_mat(n_points_final_grid,ao_num,ao_num,3))
b_mat = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i, k, ipoint, weight1, ao_i_r, ao_k_r) &
!$OMP SHARED (aos_in_r_array_transp, aos_grad_in_r_array_transp_bis, b_mat, &
!$OMP ao_num, n_points_final_grid, final_weight_at_r_vector)
!$OMP DO SCHEDULE (static)
do i = 1, ao_num
do k = 1, ao_num
do ipoint = 1, n_points_final_grid
weight1 = 0.5d0 * final_weight_at_r_vector(ipoint)
ao_i_r = aos_in_r_array_transp(ipoint,i)
ao_k_r = aos_in_r_array_transp(ipoint,k)
b_mat(ipoint,k,i,1) = weight1 * (ao_k_r * aos_grad_in_r_array_transp_bis(ipoint,i,1) - ao_i_r * aos_grad_in_r_array_transp_bis(ipoint,k,1))
b_mat(ipoint,k,i,2) = weight1 * (ao_k_r * aos_grad_in_r_array_transp_bis(ipoint,i,2) - ao_i_r * aos_grad_in_r_array_transp_bis(ipoint,k,2))
b_mat(ipoint,k,i,3) = weight1 * (ao_k_r * aos_grad_in_r_array_transp_bis(ipoint,i,3) - ao_i_r * aos_grad_in_r_array_transp_bis(ipoint,k,3))
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
tc_grad_and_lapl_ao = 0.d0
do m = 1, 3
call dgemm( "N", "N", ao_num*ao_num, ao_num*ao_num, n_points_final_grid, 1.d0 &
, int2_grad1_u12_ao(1,1,1,m), ao_num*ao_num, b_mat(1,1,1,m), n_points_final_grid &
, 1.d0, tc_grad_and_lapl_ao, ao_num*ao_num)
enddo
deallocate(b_mat)
call sum_A_At(tc_grad_and_lapl_ao(1,1,1,1), ao_num*ao_num)
endif
if(write_tc_integ.and.mpi_master) then
open(unit=11, form="unformatted", file=trim(ezfio_filename)//'/work/tc_grad_and_lapl_ao', action="write")
call ezfio_set_work_empty(.False.)
write(11) tc_grad_and_lapl_ao
close(11)
call ezfio_set_tc_keywords_io_tc_integ('Read')
endif
call wall_time(time1)
print*, ' Wall time for tc_grad_and_lapl_ao = ', time1 - time0
call print_memory_usage()
END_PROVIDER
! ---