10
0
mirror of https://github.com/QuantumPackage/qp2.git synced 2024-12-22 12:23:43 +01:00

Merge branch 'AbdAmmar-dev-stable' into dev-stable

This commit is contained in:
Anthony Scemama 2024-05-06 17:32:56 +02:00
commit 310872e177
58 changed files with 1360 additions and 7998 deletions

2
external/irpf90 vendored

@ -1 +1 @@
Subproject commit 4ab1b175fc7ed0d96c1912f13dc53579b24157a6
Subproject commit beac615343f421bd6c0571a408ba389a6d5a32ac

View File

@ -322,6 +322,12 @@ END_PROVIDER
BEGIN_PROVIDER [double precision, noL_0e]
BEGIN_DOC
!
! < Phi_left | L | Phi_right >
!
END_DOC
implicit none
integer :: i, j, k, ipoint
double precision :: t0, t1
@ -330,10 +336,6 @@ BEGIN_PROVIDER [double precision, noL_0e]
double precision, allocatable :: tmp_M(:,:), tmp_S(:), tmp_O(:), tmp_J(:,:)
double precision, allocatable :: tmp_M_priv(:,:), tmp_S_priv(:), tmp_O_priv(:), tmp_J_priv(:,:)
call wall_time(t0)
print*, " Providing noL_0e ..."
if(elec_alpha_num .eq. elec_beta_num) then
allocate(tmp(elec_beta_num))
@ -708,11 +710,6 @@ BEGIN_PROVIDER [double precision, noL_0e]
endif
call wall_time(t1)
print*, " Wall time for noL_0e (min) = ", (t1 - t0)/60.d0
print*, " noL_0e = ", noL_0e
END_PROVIDER
! ---

View File

@ -123,7 +123,7 @@ subroutine give_integrals_3_body_bi_ort_spin( n, sigma_n, l, sigma_l, k, sigma_k
endif
return
end subroutine give_integrals_3_body_bi_ort_spin
end
! ---

View File

@ -132,6 +132,7 @@ subroutine get_j1e_coef_fit_ao2(dim_fit, coef_fit)
double precision, allocatable :: A(:,:,:,:), b(:), A_tmp(:,:,:,:)
double precision, allocatable :: Pa(:,:), Pb(:,:), Pt(:,:)
double precision, allocatable :: u1e_tmp(:), tmp(:,:,:)
double precision, allocatable :: tmp1(:,:,:), tmp2(:,:,:)
double precision, allocatable :: U(:,:), D(:), Vt(:,:), work(:)
@ -176,26 +177,27 @@ subroutine get_j1e_coef_fit_ao2(dim_fit, coef_fit)
! --- --- ---
! get A
allocate(tmp(n_points_final_grid,ao_num,ao_num))
allocate(tmp1(n_points_final_grid,ao_num,ao_num), tmp2(n_points_final_grid,ao_num,ao_num))
allocate(A(ao_num,ao_num,ao_num,ao_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i, j, ipoint) &
!$OMP SHARED (n_points_final_grid, ao_num, final_weight_at_r_vector, aos_in_r_array_transp, tmp)
!$OMP SHARED (n_points_final_grid, ao_num, final_weight_at_r_vector, aos_in_r_array_transp, tmp1, tmp2)
!$OMP DO COLLAPSE(2)
do j = 1, ao_num
do i = 1, ao_num
do ipoint = 1, n_points_final_grid
tmp(ipoint,i,j) = dsqrt(final_weight_at_r_vector(ipoint)) * aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
tmp1(ipoint,i,j) = final_weight_at_r_vector(ipoint) * aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
tmp2(ipoint,i,j) = aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
call dgemm( "T", "N", ao_num*ao_num, ao_num*ao_num, n_points_final_grid, 1.d0 &
, tmp(1,1,1), n_points_final_grid, tmp(1,1,1), n_points_final_grid &
call dgemm( "T", "N", ao_num*ao_num, ao_num*ao_num, n_points_final_grid, 1.d0 &
, tmp1(1,1,1), n_points_final_grid, tmp2(1,1,1), n_points_final_grid &
, 0.d0, A(1,1,1,1), ao_num*ao_num)
allocate(A_tmp(ao_num,ao_num,ao_num,ao_num))
@ -207,13 +209,13 @@ subroutine get_j1e_coef_fit_ao2(dim_fit, coef_fit)
allocate(b(ao_num*ao_num))
do ipoint = 1, n_points_final_grid
u1e_tmp(ipoint) = dsqrt(final_weight_at_r_vector(ipoint)) * u1e_tmp(ipoint)
u1e_tmp(ipoint) = u1e_tmp(ipoint)
enddo
call dgemv("T", n_points_final_grid, ao_num*ao_num, 1.d0, tmp(1,1,1), n_points_final_grid, u1e_tmp(1), 1, 0.d0, b(1), 1)
call dgemv("T", n_points_final_grid, ao_num*ao_num, 1.d0, tmp1(1,1,1), n_points_final_grid, u1e_tmp(1), 1, 0.d0, b(1), 1)
deallocate(u1e_tmp)
deallocate(tmp)
deallocate(tmp1, tmp2)
! --- --- ---
! solve Ax = b

View File

@ -4,7 +4,7 @@
subroutine get_grad1_u12_withsq_r1_seq(ipoint, n_grid2, resx, resy, resz, res)
BEGIN_DOC
!
!
! grad_1 u(r1,r2)
!
! we use grid for r1 and extra_grid for r2
@ -167,7 +167,7 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
integer :: jpoint
integer :: i_nucl, p, mpA, npA, opA
double precision :: r2(3)
double precision :: dx, dy, dz, r12, tmp, r12_inv
double precision :: dx, dy, dz, r12, tmp
double precision :: mu_val, mu_tmp, mu_der(3)
double precision :: rn(3), f1A, grad1_f1A(3), f2A, grad2_f2A(3), g12, grad1_g12(3)
double precision :: tmp1, tmp2
@ -181,7 +181,7 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
! d/dy1 j(mu,r12) = 0.5 * [(1 - erf(mu * r12)) / r12] * (y1 - y2)
! d/dz1 j(mu,r12) = 0.5 * [(1 - erf(mu * r12)) / r12] * (z1 - z2)
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
@ -191,19 +191,15 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dx * dx + dy * dy + dz * dz
if(r12 .lt. 1d-20) then
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
r12 = dsqrt(dx * dx + dy * dy + dz * dz)
if(r12 .lt. 1d-10) then
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
cycle
endif
r12_inv = 1.d0/dsqrt(r12)
r12 = r12*r12_inv
tmp = 0.5d0 * (1.d0 - derf(mu_erf * r12)) * r12_inv
tmp = 0.5d0 * (1.d0 - derf(mu_erf * r12)) / r12
gradx(jpoint) = tmp * dx
grady(jpoint) = tmp * dy
@ -212,10 +208,10 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
elseif(j2e_type .eq. "Mur") then
! d/dx1 j(mu(r1,r2),r12) = exp(-(mu(r1,r2)*r12)**2) /(2 *sqrt(pi) * mu(r1,r2)**2 ) d/dx1 mu(r1,r2)
! d/dx1 j(mu(r1,r2),r12) = exp(-(mu(r1,r2)*r12)**2) /(2 *sqrt(pi) * mu(r1,r2)**2 ) d/dx1 mu(r1,r2)
! + 0.5 * (1 - erf(mu(r1,r2) *r12))/r12 * (x1 - x2)
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
@ -224,29 +220,23 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dsqrt(dx * dx + dy * dy + dz * dz)
r12 = dx * dx + dy * dy + dz * dz
call mu_r_val_and_grad(r1, r2, mu_val, mu_der)
mu_tmp = mu_val * r12
tmp = inv_sq_pi_2 * dexp(-mu_tmp*mu_tmp) / (mu_val * mu_val)
gradx(jpoint) = tmp * mu_der(1)
grady(jpoint) = tmp * mu_der(2)
gradz(jpoint) = tmp * mu_der(3)
if(r12 .lt. 1d-20) then
if(r12 .lt. 1d-10) then
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
cycle
endif
r12_inv = 1.d0/dsqrt(r12)
r12 = r12*r12_inv
call mu_r_val_and_grad(r1, r2, mu_val, mu_der)
mu_tmp = mu_val * r12
tmp = inv_sq_pi_2 * dexp(-mu_tmp*mu_tmp) / (mu_val * mu_val)
gradx(jpoint) = tmp * mu_der(1)
grady(jpoint) = tmp * mu_der(2)
gradz(jpoint) = tmp * mu_der(3)
tmp = 0.5d0 * (1.d0 - derf(mu_tmp)) * r12_inv
tmp = 0.5d0 * (1.d0 - derf(mu_tmp)) / r12
gradx(jpoint) = gradx(jpoint) + tmp * dx
grady(jpoint) = grady(jpoint) + tmp * dy
@ -264,7 +254,7 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
PROVIDE a_boys
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
@ -273,17 +263,14 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dx * dx + dy * dy + dz * dz
r12 = dsqrt(dx * dx + dy * dy + dz * dz)
if(r12 .lt. 1d-10) then
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
cycle
endif
r12 = dsqrt(r12)
tmp = 1.d0 + a_boys * r12
tmp = 0.5d0 / (r12 * tmp * tmp)
@ -294,13 +281,16 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
elseif(j2e_type .eq. "Boys_Handy") then
integer :: powmax
powmax = max(maxval(jBH_m),maxval(jBH_n))
integer :: powmax1, powmax, powmax2
double precision, allocatable :: f1A_power(:), f2A_power(:), double_p(:), g12_power(:)
allocate (f1A_power(-1:powmax), f2A_power(-1:powmax), g12_power(-1:powmax), double_p(0:powmax))
do p=0,powmax
powmax1 = max(maxval(jBH_m), maxval(jBH_n))
powmax2 = maxval(jBH_o)
powmax = max(powmax1, powmax2)
allocate(f1A_power(-1:powmax), f2A_power(-1:powmax), g12_power(-1:powmax), double_p(0:powmax))
do p = 0, powmax
double_p(p) = dble(p)
enddo
@ -318,11 +308,10 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
r2(2) = final_grid_points_extra(2,jpoint)
r2(3) = final_grid_points_extra(3,jpoint)
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
do i_nucl = 1, nucl_num
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
do i_nucl = 1, nucl_num
rn(1) = nucl_coord(i_nucl,1)
rn(2) = nucl_coord(i_nucl,2)
@ -332,23 +321,15 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
call jBH_elem_fct_grad(jBH_en(i_nucl), r2, rn, f2A, grad2_f2A)
call jBH_elem_fct_grad(jBH_ee(i_nucl), r1, r2, g12, grad1_g12)
! Compute powers of f1A and f2A
do p = 1, maxval(jBH_m(:,i_nucl))
do p = 1, powmax1
f1A_power(p) = f1A_power(p-1) * f1A
enddo
do p = 1, maxval(jBH_n(:,i_nucl))
f2A_power(p) = f2A_power(p-1) * f2A
enddo
do p = 1, maxval(jBH_o(:,i_nucl))
do p = 1, powmax2
g12_power(p) = g12_power(p-1) * g12
enddo
do p = 1, jBH_size
mpA = jBH_m(p,i_nucl)
npA = jBH_n(p,i_nucl)
@ -358,27 +339,22 @@ subroutine grad1_j12_r1_seq(r1, n_grid2, gradx, grady, gradz)
tmp = tmp * 0.5d0
endif
!TODO : Powers to optimize here
! tmp1 = 0.d0
! if(mpA .gt. 0) then
! tmp1 = tmp1 + dble(mpA) * f1A**(mpA-1) * f2A**npA
! endif
! if(npA .gt. 0) then
! tmp1 = tmp1 + dble(npA) * f1A**(npA-1) * f2A**mpA
! endif
! tmp1 = tmp1 * g12**(opA)
!
! tmp2 = 0.d0
! if(opA .gt. 0) then
! tmp2 = tmp2 + dble(opA) * g12**(opA-1) * (f1A**(mpA) * f2A**(npA) + f1A**(npA) * f2A**(mpA))
! endif
tmp1 = double_p(mpA) * f1A_power(mpA-1) * f2A_power(npA) + double_p(npA) * f1A_power(npA-1) * f2A_power(mpA)
tmp1 = tmp1 * g12_power(opA)
tmp2 = double_p(opA) * g12_power(opA-1) * (f1A_power(mpA) * f2A_power(npA) + f1A_power(npA) * f2A_power(mpA))
!tmp1 = 0.d0
!if(mpA .gt. 0) then
! tmp1 = tmp1 + dble(mpA) * f1A**dble(mpA-1) * f2A**dble(npA)
!endif
!if(npA .gt. 0) then
! tmp1 = tmp1 + dble(npA) * f1A**dble(npA-1) * f2A**dble(mpA)
!endif
!tmp1 = tmp1 * g12**dble(opA)
!tmp2 = 0.d0
!if(opA .gt. 0) then
! tmp2 = tmp2 + dble(opA) * g12**dble(opA-1) * (f1A**dble(mpA) * f2A**dble(npA) + f1A**dble(npA) * f2A**dble(mpA))
!endif
gradx(jpoint) = gradx(jpoint) + tmp * (tmp1 * grad1_f1A(1) + tmp2 * grad1_g12(1))
grady(jpoint) = grady(jpoint) + tmp * (tmp1 * grad1_f1A(2) + tmp2 * grad1_g12(2))
@ -418,10 +394,10 @@ subroutine grad1_jmu_r1_seq(mu, r1, n_grid2, gradx, grady, gradz)
integer :: jpoint
double precision :: r2(3)
double precision :: dx, dy, dz, r12, r12_inv, tmp
double precision :: dx, dy, dz, r12, tmp
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
@ -431,19 +407,15 @@ subroutine grad1_jmu_r1_seq(mu, r1, n_grid2, gradx, grady, gradz)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dx * dx + dy * dy + dz * dz
if(r12 .lt. 1d-20) then
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
r12 = dsqrt(dx * dx + dy * dy + dz * dz)
if(r12 .lt. 1d-10) then
gradx(jpoint) = 0.d0
grady(jpoint) = 0.d0
gradz(jpoint) = 0.d0
cycle
endif
r12_inv = 1.d0 / dsqrt(r12)
r12 = r12 * r12_inv
tmp = 0.5d0 * (1.d0 - derf(mu * r12)) * r12_inv
tmp = 0.5d0 * (1.d0 - derf(mu * r12)) / r12
gradx(jpoint) = tmp * dx
grady(jpoint) = tmp * dy
@ -467,7 +439,7 @@ subroutine j12_r1_seq(r1, n_grid2, res)
integer :: jpoint
double precision :: r2(3)
double precision :: dx, dy, dz
double precision :: mu_tmp, r12, mu_erf_inv
double precision :: mu_tmp, r12
PROVIDE final_grid_points_extra
@ -475,21 +447,20 @@ subroutine j12_r1_seq(r1, n_grid2, res)
PROVIDE mu_erf
mu_erf_inv = 1.d0 / mu_erf
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
r2(3) = final_grid_points_extra(3,jpoint)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dsqrt(dx * dx + dy * dy + dz * dz)
mu_tmp = mu_erf * r12
res(jpoint) = 0.5d0 * r12 * (1.d0 - derf(mu_tmp)) - inv_sq_pi_2 * dexp(-mu_tmp*mu_tmp) * mu_erf_inv
res(jpoint) = 0.5d0 * r12 * (1.d0 - derf(mu_tmp)) - inv_sq_pi_2 * dexp(-mu_tmp*mu_tmp) / mu_erf
enddo
elseif(j2e_type .eq. "Boys") then
@ -498,7 +469,7 @@ subroutine j12_r1_seq(r1, n_grid2, res)
PROVIDE a_boys
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
@ -540,19 +511,19 @@ subroutine jmu_r1_seq(mu, r1, n_grid2, res)
tmp1 = inv_sq_pi_2 / mu
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r2(1) = final_grid_points_extra(1,jpoint)
r2(2) = final_grid_points_extra(2,jpoint)
r2(3) = final_grid_points_extra(3,jpoint)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dsqrt(dx * dx + dy * dy + dz * dz)
tmp2 = mu * r12
res(jpoint) = 0.5d0 * r12 * (1.d0 - derf(tmp2)) - tmp1 * dexp(-tmp2*tmp2)
enddo
@ -579,7 +550,7 @@ subroutine env_nucl_r1_seq(n_grid2, res)
res = 1.d0
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r(1) = final_grid_points_extra(1,jpoint)
r(2) = final_grid_points_extra(2,jpoint)
r(3) = final_grid_points_extra(3,jpoint)
@ -598,7 +569,7 @@ subroutine env_nucl_r1_seq(n_grid2, res)
res = 1.d0
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r(1) = final_grid_points_extra(1,jpoint)
r(2) = final_grid_points_extra(2,jpoint)
r(3) = final_grid_points_extra(3,jpoint)
@ -618,7 +589,7 @@ subroutine env_nucl_r1_seq(n_grid2, res)
res = 1.d0
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r(1) = final_grid_points_extra(1,jpoint)
r(2) = final_grid_points_extra(2,jpoint)
r(3) = final_grid_points_extra(3,jpoint)
@ -636,7 +607,7 @@ subroutine env_nucl_r1_seq(n_grid2, res)
res = 1.d0
do jpoint = 1, n_points_extra_final_grid ! r2
do jpoint = 1, n_points_extra_final_grid ! r2
r(1) = final_grid_points_extra(1,jpoint)
r(2) = final_grid_points_extra(2,jpoint)
r(3) = final_grid_points_extra(3,jpoint)
@ -666,7 +637,7 @@ end
subroutine get_grad1_u12_2e_r1_seq(ipoint, n_grid2, resx, resy, resz)
BEGIN_DOC
!
!
! grad_1 u_2e(r1,r2)
!
! we use grid for r1 and extra_grid for r2
@ -786,7 +757,7 @@ end
subroutine get_u12_2e_r1_seq(ipoint, n_grid2, res)
BEGIN_DOC
!
!
! u_2e(r1,r2)
!
! we use grid for r1 and extra_grid for r2
@ -909,7 +880,7 @@ subroutine jBH_elem_fct_grad(alpha, r1, r2, fct, grad1_fct)
endif
return
end
end
! ---

View File

@ -179,7 +179,7 @@ double precision function num_v_ij_erf_rk_cst_mu_env(i, j, ipoint)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dsqrt( dx * dx + dy * dy + dz * dz )
r12 = dsqrt(dx*dx + dy*dy + dz*dz)
if(r12 .lt. 1d-10) cycle
tmp1 = (derf(mu_erf * r12) - 1.d0) / r12
@ -228,7 +228,7 @@ subroutine num_x_v_ij_erf_rk_cst_mu_env(i, j, ipoint, integ)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dsqrt( dx * dx + dy * dy + dz * dz )
r12 = dsqrt(dx*dx + dy*dy + dz*dz)
if(r12 .lt. 1d-10) cycle
tmp1 = (derf(mu_erf * r12) - 1.d0) / r12
@ -530,7 +530,7 @@ subroutine num_int2_u_grad1u_total_env2(i, j, ipoint, integ)
dx = r1(1) - r2(1)
dy = r1(2) - r2(2)
dz = r1(3) - r2(3)
r12 = dsqrt( dx * dx + dy * dy + dz * dz )
r12 = dsqrt(dx*dx + dy*dy + dz*dz)
if(r12 .lt. 1d-10) cycle
tmp0 = env_nucl(r2)

View File

@ -63,12 +63,10 @@
do i_pass = 1, n_pass
ii = (i_pass-1)*n_blocks + 1
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i_blocks, ipoint) &
!$OMP SHARED (n_blocks, n_points_extra_final_grid, ii, &
!$OMP final_grid_points, tmp_grad1_u12, &
!$OMP tmp_grad1_u12_squared)
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i_blocks, ipoint) &
!$OMP SHARED (n_blocks, n_points_extra_final_grid, ii, final_grid_points, tmp_grad1_u12, tmp_grad1_u12_squared)
!$OMP DO
do i_blocks = 1, n_blocks
ipoint = ii - 1 + i_blocks ! r1
@ -99,12 +97,10 @@
ii = n_pass*n_blocks + 1
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i_rest, ipoint) &
!$OMP SHARED (n_rest, n_points_extra_final_grid, ii, &
!$OMP final_grid_points, tmp_grad1_u12, &
!$OMP tmp_grad1_u12_squared)
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i_rest, ipoint) &
!$OMP SHARED (n_rest, n_points_extra_final_grid, ii, final_grid_points, tmp_grad1_u12, tmp_grad1_u12_squared)
!$OMP DO
do i_rest = 1, n_rest
ipoint = ii - 1 + i_rest ! r1

View File

@ -1125,6 +1125,7 @@ subroutine test_fit_coef_A1()
double precision :: accu, norm, diff
double precision, allocatable :: A1(:,:)
double precision, allocatable :: A2(:,:,:,:), tmp(:,:,:)
double precision, allocatable :: tmp1(:,:,:), tmp2(:,:,:)
! ---
@ -1165,16 +1166,17 @@ subroutine test_fit_coef_A1()
call wall_time(t1)
allocate(tmp(ao_num,ao_num,n_points_final_grid))
allocate(tmp1(ao_num,ao_num,n_points_final_grid), tmp2(ao_num,ao_num,n_points_final_grid))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i, j, ipoint) &
!$OMP SHARED (n_points_final_grid, ao_num, final_weight_at_r_vector, aos_in_r_array_transp, tmp)
!$OMP SHARED (n_points_final_grid, ao_num, final_weight_at_r_vector, aos_in_r_array_transp, tmp1, tmp2)
!$OMP DO COLLAPSE(2)
do j = 1, ao_num
do i = 1, ao_num
do ipoint = 1, n_points_final_grid
tmp(i,j,ipoint) = dsqrt(final_weight_at_r_vector(ipoint)) * aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
tmp1(i,j,ipoint) = final_weight_at_r_vector(ipoint) * aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
tmp2(i,j,ipoint) = aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
enddo
enddo
enddo
@ -1184,9 +1186,9 @@ subroutine test_fit_coef_A1()
allocate(A2(ao_num,ao_num,ao_num,ao_num))
call dgemm( "N", "T", ao_num*ao_num, ao_num*ao_num, n_points_final_grid, 1.d0 &
, tmp(1,1,1), ao_num*ao_num, tmp(1,1,1), ao_num*ao_num &
, tmp1(1,1,1), ao_num*ao_num, tmp2(1,1,1), ao_num*ao_num &
, 0.d0, A2(1,1,1,1), ao_num*ao_num)
deallocate(tmp)
deallocate(tmp1, tmp2)
call wall_time(t2)
print*, ' WALL TIME FOR A2 (min) =', (t2-t1)/60.d0
@ -1238,6 +1240,7 @@ subroutine test_fit_coef_inv()
double precision, allocatable :: A1(:,:), A1_inv(:,:), A1_tmp(:,:)
double precision, allocatable :: A2(:,:,:,:), tmp(:,:,:), A2_inv(:,:,:,:)
double precision, allocatable :: U(:,:), D(:), Vt(:,:), work(:), A2_tmp(:,:,:,:)
double precision, allocatable :: tmp1(:,:,:), tmp2(:,:,:)
cutoff_svd = 5d-8
@ -1286,16 +1289,17 @@ subroutine test_fit_coef_inv()
call wall_time(t1)
allocate(tmp(n_points_final_grid,ao_num,ao_num))
allocate(tmp1(n_points_final_grid,ao_num,ao_num), tmp2(n_points_final_grid,ao_num,ao_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (i, j, ipoint) &
!$OMP SHARED (n_points_final_grid, ao_num, final_weight_at_r_vector, aos_in_r_array_transp, tmp)
!$OMP SHARED (n_points_final_grid, ao_num, final_weight_at_r_vector, aos_in_r_array_transp, tmp1, tmp2)
!$OMP DO COLLAPSE(2)
do j = 1, ao_num
do i = 1, ao_num
do ipoint = 1, n_points_final_grid
tmp(ipoint,i,j) = dsqrt(final_weight_at_r_vector(ipoint)) * aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
tmp1(ipoint,i,j) = final_weight_at_r_vector(ipoint) * aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
tmp2(ipoint,i,j) = aos_in_r_array_transp(ipoint,i) * aos_in_r_array_transp(ipoint,j)
enddo
enddo
enddo
@ -1304,11 +1308,11 @@ subroutine test_fit_coef_inv()
allocate(A2(ao_num,ao_num,ao_num,ao_num))
call dgemm( "T", "N", ao_num*ao_num, ao_num*ao_num, n_points_final_grid, 1.d0 &
, tmp(1,1,1), n_points_final_grid, tmp(1,1,1), n_points_final_grid &
call dgemm( "T", "N", ao_num*ao_num, ao_num*ao_num, n_points_final_grid, 1.d0 &
, tmp1(1,1,1), n_points_final_grid, tmp2(1,1,1), n_points_final_grid &
, 0.d0, A2(1,1,1,1), ao_num*ao_num)
deallocate(tmp)
deallocate(tmp1, tmp2)
call wall_time(t2)
print*, ' WALL TIME FOR A2 (min) =', (t2-t1)/60.d0

File diff suppressed because it is too large Load Diff

View File

@ -273,60 +273,6 @@ end
! ---
subroutine lapack_diag_non_sym_right(n, A, WR, WI, VR)
implicit none
integer, intent(in) :: n
double precision, intent(in) :: A(n,n)
double precision, intent(out) :: WR(n), WI(n), VR(n,n)
integer :: i, lda, ldvl, ldvr, LWORK, INFO
double precision, allocatable :: Atmp(:,:), WORK(:), VL(:,:)
lda = n
ldvl = 1
ldvr = n
allocate( Atmp(n,n), VL(1,1) )
Atmp(1:n,1:n) = A(1:n,1:n)
allocate(WORK(1))
LWORK = -1
call dgeev('N', 'V', n, Atmp, lda, WR, WI, VL, ldvl, VR, ldvr, WORK, LWORK, INFO)
if(INFO.gt.0)then
print*,'dgeev failed !!',INFO
stop
endif
LWORK = max(int(WORK(1)), 1) ! this is the optimal size of WORK
deallocate(WORK)
allocate(WORK(LWORK))
! Actual diagonalization
call dgeev('N', 'V', n, Atmp, lda, WR, WI, VL, ldvl, VR, ldvr, WORK, LWORK, INFO)
if(INFO.ne.0) then
print*,'dgeev failed !!', INFO
stop
endif
deallocate(Atmp, WORK, VL)
! print *, ' JOBL = F'
! print *, ' eigenvalues'
! do i = 1, n
! write(*, '(1000(F16.10,X))') WR(i), WI(i)
! enddo
! print *, ' right eigenvect'
! do i = 1, n
! write(*, '(1000(F16.10,X))') VR(:,i)
! enddo
end
! ---
subroutine non_hrmt_real_diag(n, A, leigvec, reigvec, n_real_eigv, eigval)
BEGIN_DOC
@ -1780,70 +1726,6 @@ end
! ---
subroutine check_weighted_biorthog(n, m, W, Vl, Vr, thr_d, thr_nd, accu_d, accu_nd, S, stop_ifnot)
implicit none
integer, intent(in) :: n, m
double precision, intent(in) :: Vl(n,m), Vr(n,m), W(n,n)
double precision, intent(in) :: thr_d, thr_nd
logical, intent(in) :: stop_ifnot
double precision, intent(out) :: accu_d, accu_nd, S(m,m)
integer :: i, j
double precision, allocatable :: SS(:,:), tmp(:,:)
print *, ' check weighted bi-orthogonality'
! ---
allocate(tmp(m,n))
call dgemm( 'T', 'N', m, n, n, 1.d0 &
, Vl, size(Vl, 1), W, size(W, 1) &
, 0.d0, tmp, size(tmp, 1) )
call dgemm( 'N', 'N', m, m, n, 1.d0 &
, tmp, size(tmp, 1), Vr, size(Vr, 1) &
, 0.d0, S, size(S, 1) )
deallocate(tmp)
!print *, ' overlap matrix:'
!do i = 1, m
! write(*,'(1000(F16.10,X))') S(i,:)
!enddo
accu_d = 0.d0
accu_nd = 0.d0
do i = 1, m
do j = 1, m
if(i==j) then
accu_d = accu_d + dabs(S(i,i))
else
accu_nd = accu_nd + S(j,i) * S(j,i)
endif
enddo
enddo
accu_nd = dsqrt(accu_nd)
print *, ' accu_nd = ', accu_nd
print *, ' accu_d = ', dabs(accu_d-dble(m))/dble(m)
! ---
if( stop_ifnot .and. ((accu_nd .gt. thr_nd) .or. dabs(accu_d-dble(m))/dble(m) .gt. thr_d) ) then
print *, ' non bi-orthogonal vectors !'
print *, ' accu_nd = ', accu_nd
print *, ' accu_d = ', dabs(accu_d-dble(m))/dble(m)
!print *, ' overlap matrix:'
!do i = 1, m
! write(*,'(1000(F16.10,X))') S(i,:)
!enddo
stop
endif
end
! ---
subroutine check_biorthog(n, m, Vl, Vr, accu_d, accu_nd, S, thr_d, thr_nd, stop_ifnot)
implicit none

View File

@ -1,670 +0,0 @@
subroutine non_hrmt_diag_split_degen_bi_orthog(n, A, leigvec, reigvec, n_real_eigv, eigval)
BEGIN_DOC
!
! routine which returns the sorted REAL EIGENVALUES ONLY and corresponding LEFT/RIGHT eigenvetors
!
! of a non hermitian matrix A(n,n)
!
! n_real_eigv is the number of real eigenvalues, which might be smaller than the dimension "n"
!
END_DOC
implicit none
integer, intent(in) :: n
double precision, intent(in) :: A(n,n)
integer, intent(out) :: n_real_eigv
double precision, intent(out) :: reigvec(n,n), leigvec(n,n), eigval(n)
double precision, allocatable :: reigvec_tmp(:,:), leigvec_tmp(:,:)
integer :: i, j, n_degen,k , iteration
double precision :: shift_current
double precision :: r,thr,accu_d, accu_nd
integer, allocatable :: iorder_origin(:),iorder(:)
double precision, allocatable :: WR(:), WI(:), Vl(:,:), VR(:,:),S(:,:)
double precision, allocatable :: Aw(:,:),diag_elem(:),A_save(:,:)
double precision, allocatable :: im_part(:),re_part(:)
double precision :: accu,thr_cut, thr_norm=1d0
thr_cut = 1.d-15
print*,'Computing the left/right eigenvectors ...'
print*,'Using the degeneracy splitting algorithm'
! initialization
shift_current = 1.d-15
iteration = 0
print*,'***** iteration = ',iteration
! pre-processing the matrix :: sorting by diagonal elements
allocate(reigvec_tmp(n,n), leigvec_tmp(n,n))
allocate(diag_elem(n),iorder_origin(n),A_save(n,n))
! print*,'Aw'
do i = 1, n
iorder_origin(i) = i
diag_elem(i) = A(i,i)
! write(*,'(100(F16.10,X))')A(:,i)
enddo
call dsort(diag_elem, iorder_origin, n)
do i = 1, n
do j = 1, n
A_save(j,i) = A(iorder_origin(j),iorder_origin(i))
enddo
enddo
allocate(WR(n), WI(n), VL(n,n), VR(n,n), Aw(n,n))
allocate(im_part(n),iorder(n))
allocate( S(n,n) )
Aw = A_save
call cancel_small_elmts(aw,n,thr_cut)
call lapack_diag_non_sym(n,Aw,WR,WI,VL,VR)
do i = 1, n
im_part(i) = -dabs(WI(i))
iorder(i) = i
enddo
call dsort(im_part, iorder, n)
n_real_eigv = 0
do i = 1, n
if(dabs(WI(i)).lt.1.d-20)then
n_real_eigv += 1
else
! print*,'Found an imaginary component to eigenvalue'
! print*,'Re(i) + Im(i)',WR(i),WI(i)
endif
enddo
if(n_real_eigv.ne.n)then
shift_current = max(10.d0 * dabs(im_part(1)),shift_current*10.d0)
print*,'Largest imaginary part found in eigenvalues = ',im_part(1)
print*,'Splitting the degeneracies by ',shift_current
else
print*,'All eigenvalues are real !'
endif
do while(n_real_eigv.ne.n)
iteration += 1
print*,'***** iteration = ',iteration
if(shift_current.gt.1.d-3)then
print*,'shift_current > 1.d-3 !!'
print*,'Your matrix intrinsically contains complex eigenvalues'
stop
endif
Aw = A_save
call cancel_small_elmts(Aw,n,thr_cut)
call split_matrix_degen(Aw,n,shift_current)
call lapack_diag_non_sym(n,Aw,WR,WI,VL,VR)
n_real_eigv = 0
do i = 1, n
if(dabs(WI(i)).lt.1.d-20)then
n_real_eigv+= 1
else
! print*,'Found an imaginary component to eigenvalue'
! print*,'Re(i) + Im(i)',WR(i),WI(i)
endif
enddo
if(n_real_eigv.ne.n)then
do i = 1, n
im_part(i) = -dabs(WI(i))
iorder(i) = i
enddo
call dsort(im_part, iorder, n)
shift_current = max(10.d0 * dabs(im_part(1)),shift_current*10.d0)
print*,'Largest imaginary part found in eigenvalues = ',im_part(1)
print*,'Splitting the degeneracies by ',shift_current
else
print*,'All eigenvalues are real !'
endif
enddo
!!!!!!!!!!!!!!!! SORTING THE EIGENVALUES
do i = 1, n
eigval(i) = WR(i)
iorder(i) = i
enddo
call dsort(eigval,iorder,n)
do i = 1, n
! print*,'eigval(i) = ',eigval(i)
reigvec_tmp(:,i) = VR(:,iorder(i))
leigvec_tmp(:,i) = Vl(:,iorder(i))
enddo
!!! ONCE ALL EIGENVALUES ARE REAL ::: CHECK BI-ORTHONORMALITY
! check bi-orthogonality
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print *, ' accu_nd bi-orthog = ', accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print *, ' '
print *, ' bi-orthogonality: not imposed yet'
print *, ' '
print *, ' '
print *, ' orthog between degen eigenvect'
print *, ' '
double precision, allocatable :: S_nh_inv_half(:,:)
allocate(S_nh_inv_half(n,n))
logical :: complex_root
deallocate(S_nh_inv_half)
call impose_orthog_degen_eigvec(n, eigval, reigvec_tmp)
call impose_orthog_degen_eigvec(n, eigval, leigvec_tmp)
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'New vectors not bi-orthonormals at ',accu_nd
call impose_biorthog_qr(n, n, leigvec_tmp, reigvec_tmp, S)
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'New vectors not bi-orthonormals at ',accu_nd
print*,'Must be a deep problem ...'
stop
endif
endif
endif
!! EIGENVECTORS SORTED AND BI-ORTHONORMAL
do i = 1, n
do j = 1, n
VR(iorder_origin(j),i) = reigvec_tmp(j,i)
VL(iorder_origin(j),i) = leigvec_tmp(j,i)
enddo
enddo
!! RECOMPUTING THE EIGENVALUES
eigval = 0.d0
do i = 1, n
iorder(i) = i
accu = 0.d0
do j = 1, n
accu += VL(j,i) * VR(j,i)
do k = 1, n
eigval(i) += VL(j,i) * A(j,k) * VR(k,i)
enddo
enddo
eigval(i) *= 1.d0/accu
! print*,'eigval(i) = ',eigval(i)
enddo
!! RESORT JUST TO BE SURE
call dsort(eigval, iorder, n)
do i = 1, n
do j = 1, n
reigvec(j,i) = VR(j,iorder(i))
leigvec(j,i) = VL(j,iorder(i))
enddo
enddo
print*,'Checking for final reigvec/leigvec'
shift_current = max(1.d-10,shift_current)
print*,'Thr for eigenvectors = ',shift_current
call check_EIGVEC(n, n, A, eigval, leigvec, reigvec, shift_current, thr_norm, .false.)
call check_biorthog(n, n, leigvec, reigvec, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print *, ' accu_nd bi-orthog = ', accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'Something went wrong in non_hrmt_diag_split_degen_bi_orthog'
print*,'Eigenvectors are not bi orthonormal ..'
print*,'accu_nd = ',accu_nd
stop
endif
end
subroutine non_hrmt_diag_split_degen_s_inv_half(n, A, leigvec, reigvec, n_real_eigv, eigval)
BEGIN_DOC
!
! routine which returns the sorted REAL EIGENVALUES ONLY and corresponding LEFT/RIGHT eigenvetors
!
! of a non hermitian matrix A(n,n)
!
! n_real_eigv is the number of real eigenvalues, which might be smaller than the dimension "n"
!
END_DOC
implicit none
integer, intent(in) :: n
double precision, intent(in) :: A(n,n)
integer, intent(out) :: n_real_eigv
double precision, intent(out) :: reigvec(n,n), leigvec(n,n), eigval(n)
double precision, allocatable :: reigvec_tmp(:,:), leigvec_tmp(:,:)
integer :: i, j, n_degen,k , iteration
double precision :: shift_current
double precision :: r,thr,accu_d, accu_nd
integer, allocatable :: iorder_origin(:),iorder(:)
double precision, allocatable :: WR(:), WI(:), Vl(:,:), VR(:,:),S(:,:)
double precision, allocatable :: Aw(:,:),diag_elem(:),A_save(:,:)
double precision, allocatable :: im_part(:),re_part(:)
double precision :: accu,thr_cut, thr_norm=1.d0
double precision, allocatable :: S_nh_inv_half(:,:)
logical :: complex_root
thr_cut = 1.d-15
print*,'Computing the left/right eigenvectors ...'
print*,'Using the degeneracy splitting algorithm'
! initialization
shift_current = 1.d-15
iteration = 0
print*,'***** iteration = ',iteration
! pre-processing the matrix :: sorting by diagonal elements
allocate(reigvec_tmp(n,n), leigvec_tmp(n,n))
allocate(diag_elem(n),iorder_origin(n),A_save(n,n))
! print*,'Aw'
do i = 1, n
iorder_origin(i) = i
diag_elem(i) = A(i,i)
! write(*,'(100(F16.10,X))')A(:,i)
enddo
call dsort(diag_elem, iorder_origin, n)
do i = 1, n
do j = 1, n
A_save(j,i) = A(iorder_origin(j),iorder_origin(i))
enddo
enddo
allocate(WR(n), WI(n), VL(n,n), VR(n,n), Aw(n,n))
allocate(im_part(n),iorder(n))
allocate( S(n,n) )
allocate(S_nh_inv_half(n,n))
Aw = A_save
call cancel_small_elmts(aw,n,thr_cut)
call lapack_diag_non_sym(n,Aw,WR,WI,VL,VR)
do i = 1, n
im_part(i) = -dabs(WI(i))
iorder(i) = i
enddo
call dsort(im_part, iorder, n)
n_real_eigv = 0
do i = 1, n
if(dabs(WI(i)).lt.1.d-20)then
n_real_eigv += 1
else
! print*,'Found an imaginary component to eigenvalue'
! print*,'Re(i) + Im(i)',WR(i),WI(i)
endif
enddo
if(n_real_eigv.ne.n)then
shift_current = max(10.d0 * dabs(im_part(1)),shift_current*10.d0)
print*,'Largest imaginary part found in eigenvalues = ',im_part(1)
print*,'Splitting the degeneracies by ',shift_current
else
print*,'All eigenvalues are real !'
endif
do while(n_real_eigv.ne.n)
iteration += 1
print*,'***** iteration = ',iteration
if(shift_current.gt.1.d-3)then
print*,'shift_current > 1.d-3 !!'
print*,'Your matrix intrinsically contains complex eigenvalues'
stop
endif
Aw = A_save
! thr_cut = shift_current
call cancel_small_elmts(Aw,n,thr_cut)
call split_matrix_degen(Aw,n,shift_current)
call lapack_diag_non_sym(n,Aw,WR,WI,VL,VR)
n_real_eigv = 0
do i = 1, n
if(dabs(WI(i)).lt.1.d-20)then
n_real_eigv+= 1
else
! print*,'Found an imaginary component to eigenvalue'
! print*,'Re(i) + Im(i)',WR(i),WI(i)
endif
enddo
if(n_real_eigv.ne.n)then
do i = 1, n
im_part(i) = -dabs(WI(i))
iorder(i) = i
enddo
call dsort(im_part, iorder, n)
shift_current = max(10.d0 * dabs(im_part(1)),shift_current*10.d0)
print*,'Largest imaginary part found in eigenvalues = ',im_part(1)
print*,'Splitting the degeneracies by ',shift_current
else
print*,'All eigenvalues are real !'
endif
enddo
!!!!!!!!!!!!!!!! SORTING THE EIGENVALUES
do i = 1, n
eigval(i) = WR(i)
iorder(i) = i
enddo
call dsort(eigval,iorder,n)
do i = 1, n
! print*,'eigval(i) = ',eigval(i)
reigvec_tmp(:,i) = VR(:,iorder(i))
leigvec_tmp(:,i) = Vl(:,iorder(i))
enddo
!!! ONCE ALL EIGENVALUES ARE REAL ::: CHECK BI-ORTHONORMALITY
! check bi-orthogonality
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print *, ' accu_nd bi-orthog = ', accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print *, ' '
print *, ' bi-orthogonality: not imposed yet'
if(complex_root) then
print *, ' '
print *, ' '
print *, ' orthog between degen eigenvect'
print *, ' '
! bi-orthonormalization using orthogonalization of left, right and then QR between left and right
call impose_orthog_degen_eigvec(n, eigval, reigvec_tmp) ! orthogonalization of reigvec
call impose_orthog_degen_eigvec(n, eigval, leigvec_tmp) ! orthogonalization of leigvec
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'New vectors not bi-orthonormals at ', accu_nd
call get_inv_half_nonsymmat_diago(S, n, S_nh_inv_half, complex_root)
if(complex_root)then
call impose_biorthog_qr(n, n, leigvec_tmp, reigvec_tmp, S) ! bi-orthonormalization using QR
else
print*,'S^{-1/2} exists !!'
call bi_ortho_s_inv_half(n,leigvec_tmp,reigvec_tmp,S_nh_inv_half) ! use of S^{-1/2} bi-orthonormalization
endif
endif
else ! the matrix S^{-1/2} exists
print*,'S^{-1/2} exists !!'
call bi_ortho_s_inv_half(n,leigvec_tmp,reigvec_tmp,S_nh_inv_half) ! use of S^{-1/2} bi-orthonormalization
endif
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'New vectors not bi-orthonormals at ',accu_nd
print*,'Must be a deep problem ...'
stop
endif
endif
!! EIGENVECTORS SORTED AND BI-ORTHONORMAL
do i = 1, n
do j = 1, n
VR(iorder_origin(j),i) = reigvec_tmp(j,i)
VL(iorder_origin(j),i) = leigvec_tmp(j,i)
enddo
enddo
!! RECOMPUTING THE EIGENVALUES
eigval = 0.d0
do i = 1, n
iorder(i) = i
accu = 0.d0
do j = 1, n
accu += VL(j,i) * VR(j,i)
do k = 1, n
eigval(i) += VL(j,i) * A(j,k) * VR(k,i)
enddo
enddo
eigval(i) *= 1.d0/accu
! print*,'eigval(i) = ',eigval(i)
enddo
!! RESORT JUST TO BE SURE
call dsort(eigval, iorder, n)
do i = 1, n
do j = 1, n
reigvec(j,i) = VR(j,iorder(i))
leigvec(j,i) = VL(j,iorder(i))
enddo
enddo
print*,'Checking for final reigvec/leigvec'
shift_current = max(1.d-10,shift_current)
print*,'Thr for eigenvectors = ',shift_current
call check_EIGVEC(n, n, A, eigval, leigvec, reigvec, shift_current, thr_norm, .false.)
call check_biorthog(n, n, leigvec, reigvec, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print *, ' accu_nd bi-orthog = ', accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'Something went wrong in non_hrmt_diag_split_degen_bi_orthog'
print*,'Eigenvectors are not bi orthonormal ..'
print*,'accu_nd = ',accu_nd
stop
endif
end
subroutine non_hrmt_fock_mat(n, A, leigvec, reigvec, n_real_eigv, eigval)
BEGIN_DOC
!
! routine returning the eigenvalues and left/right eigenvectors of the TC fock matrix
!
END_DOC
implicit none
integer, intent(in) :: n
double precision, intent(in) :: A(n,n)
integer, intent(out) :: n_real_eigv
double precision, intent(out) :: reigvec(n,n), leigvec(n,n), eigval(n)
double precision, allocatable :: reigvec_tmp(:,:), leigvec_tmp(:,:)
integer :: i, j, n_degen,k , iteration
double precision :: shift_current
double precision :: r,thr,accu_d, accu_nd
integer, allocatable :: iorder_origin(:),iorder(:)
double precision, allocatable :: WR(:), WI(:), Vl(:,:), VR(:,:),S(:,:)
double precision, allocatable :: Aw(:,:),diag_elem(:),A_save(:,:)
double precision, allocatable :: im_part(:),re_part(:)
double precision :: accu,thr_cut
double precision, allocatable :: S_nh_inv_half(:,:)
logical :: complex_root
double precision :: thr_norm=1d0
thr_cut = 1.d-15
print*,'Computing the left/right eigenvectors ...'
print*,'Using the degeneracy splitting algorithm'
! initialization
shift_current = 1.d-15
iteration = 0
print*,'***** iteration = ',iteration
! pre-processing the matrix :: sorting by diagonal elements
allocate(reigvec_tmp(n,n), leigvec_tmp(n,n))
allocate(diag_elem(n),iorder_origin(n),A_save(n,n))
! print*,'Aw'
do i = 1, n
iorder_origin(i) = i
diag_elem(i) = A(i,i)
! write(*,'(100(F16.10,X))')A(:,i)
enddo
call dsort(diag_elem, iorder_origin, n)
do i = 1, n
do j = 1, n
A_save(j,i) = A(iorder_origin(j),iorder_origin(i))
enddo
enddo
allocate(WR(n), WI(n), VL(n,n), VR(n,n), Aw(n,n))
allocate(im_part(n),iorder(n))
allocate( S(n,n) )
allocate(S_nh_inv_half(n,n))
Aw = A_save
call cancel_small_elmts(aw,n,thr_cut)
call lapack_diag_non_sym(n,Aw,WR,WI,VL,VR)
do i = 1, n
im_part(i) = -dabs(WI(i))
iorder(i) = i
enddo
call dsort(im_part, iorder, n)
n_real_eigv = 0
do i = 1, n
if(dabs(WI(i)).lt.1.d-20)then
n_real_eigv += 1
else
! print*,'Found an imaginary component to eigenvalue'
! print*,'Re(i) + Im(i)',WR(i),WI(i)
endif
enddo
if(n_real_eigv.ne.n)then
shift_current = max(10.d0 * dabs(im_part(1)),shift_current*10.d0)
print*,'Largest imaginary part found in eigenvalues = ',im_part(1)
print*,'Splitting the degeneracies by ',shift_current
else
print*,'All eigenvalues are real !'
endif
do while(n_real_eigv.ne.n)
iteration += 1
print*,'***** iteration = ',iteration
if(shift_current.gt.1.d-3)then
print*,'shift_current > 1.d-3 !!'
print*,'Your matrix intrinsically contains complex eigenvalues'
stop
endif
Aw = A_save
! thr_cut = shift_current
call cancel_small_elmts(Aw,n,thr_cut)
call split_matrix_degen(Aw,n,shift_current)
call lapack_diag_non_sym(n,Aw,WR,WI,VL,VR)
n_real_eigv = 0
do i = 1, n
if(dabs(WI(i)).lt.1.d-20)then
n_real_eigv+= 1
else
! print*,'Found an imaginary component to eigenvalue'
! print*,'Re(i) + Im(i)',WR(i),WI(i)
endif
enddo
if(n_real_eigv.ne.n)then
do i = 1, n
im_part(i) = -dabs(WI(i))
iorder(i) = i
enddo
call dsort(im_part, iorder, n)
shift_current = max(10.d0 * dabs(im_part(1)),shift_current*10.d0)
print*,'Largest imaginary part found in eigenvalues = ',im_part(1)
print*,'Splitting the degeneracies by ',shift_current
else
print*,'All eigenvalues are real !'
endif
enddo
!!!!!!!!!!!!!!!! SORTING THE EIGENVALUES
do i = 1, n
eigval(i) = WR(i)
iorder(i) = i
enddo
call dsort(eigval,iorder,n)
do i = 1, n
! print*,'eigval(i) = ',eigval(i)
reigvec_tmp(:,i) = VR(:,iorder(i))
leigvec_tmp(:,i) = Vl(:,iorder(i))
enddo
!!! ONCE ALL EIGENVALUES ARE REAL ::: CHECK BI-ORTHONORMALITY
! check bi-orthogonality
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print *, ' accu_nd bi-orthog = ', accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print *, ' '
print *, ' bi-orthogonality: not imposed yet'
print *, ' '
print *, ' '
print *, ' Using impose_unique_biorthog_degen_eigvec'
print *, ' '
! bi-orthonormalization using orthogonalization of left, right and then QR between left and right
call impose_unique_biorthog_degen_eigvec(n, eigval, mo_coef, leigvec_tmp, reigvec_tmp)
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print*,'accu_nd = ',accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'New vectors not bi-orthonormals at ',accu_nd
call get_inv_half_nonsymmat_diago(S, n, S_nh_inv_half,complex_root)
if(complex_root)then
print*,'S^{-1/2} does not exits, using QR bi-orthogonalization'
call impose_biorthog_qr(n, n, leigvec_tmp, reigvec_tmp, S) ! bi-orthonormalization using QR
else
print*,'S^{-1/2} exists !!'
call bi_ortho_s_inv_half(n,leigvec_tmp,reigvec_tmp,S_nh_inv_half) ! use of S^{-1/2} bi-orthonormalization
endif
endif
call check_biorthog(n, n, leigvec_tmp, reigvec_tmp, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'New vectors not bi-orthonormals at ',accu_nd
print*,'Must be a deep problem ...'
stop
endif
endif
!! EIGENVECTORS SORTED AND BI-ORTHONORMAL
do i = 1, n
do j = 1, n
VR(iorder_origin(j),i) = reigvec_tmp(j,i)
VL(iorder_origin(j),i) = leigvec_tmp(j,i)
enddo
enddo
!! RECOMPUTING THE EIGENVALUES
eigval = 0.d0
do i = 1, n
iorder(i) = i
accu = 0.d0
do j = 1, n
accu += VL(j,i) * VR(j,i)
do k = 1, n
eigval(i) += VL(j,i) * A(j,k) * VR(k,i)
enddo
enddo
eigval(i) *= 1.d0/accu
! print*,'eigval(i) = ',eigval(i)
enddo
!! RESORT JUST TO BE SURE
call dsort(eigval, iorder, n)
do i = 1, n
do j = 1, n
reigvec(j,i) = VR(j,iorder(i))
leigvec(j,i) = VL(j,iorder(i))
enddo
enddo
print*,'Checking for final reigvec/leigvec'
shift_current = max(1.d-10,shift_current)
print*,'Thr for eigenvectors = ',shift_current
call check_EIGVEC(n, n, A, eigval, leigvec, reigvec, shift_current, thr_norm, .false.)
call check_biorthog(n, n, leigvec, reigvec, accu_d, accu_nd, S, thresh_biorthog_diag, thresh_biorthog_nondiag, .false.)
print *, ' accu_nd bi-orthog = ', accu_nd
if(accu_nd .lt. thresh_biorthog_nondiag) then
print *, ' bi-orthogonality: ok'
else
print*,'Something went wrong in non_hrmt_diag_split_degen_bi_orthog'
print*,'Eigenvectors are not bi orthonormal ..'
print*,'accu_nd = ',accu_nd
stop
endif
end

View File

@ -183,11 +183,3 @@ BEGIN_PROVIDER [ double precision, x_W_ij_erf_rk, ( n_points_final_grid,3,mo_num
END_PROVIDER
BEGIN_PROVIDER [ double precision, sqrt_weight_at_r, (n_points_final_grid)]
implicit none
integer :: ipoint
do ipoint = 1, n_points_final_grid
sqrt_weight_at_r(ipoint) = dsqrt(final_weight_at_r_vector(ipoint))
enddo
END_PROVIDER

View File

@ -5,3 +5,4 @@ bi_ortho_mos
tc_keywords
non_hermit_dav
dav_general_mat
tc_scf

View File

@ -22,6 +22,7 @@ BEGIN_PROVIDER [double precision, htilde_matrix_elmt_bi_ortho, (N_det,N_det)]
if(noL_standard) then
PROVIDE noL_0e
print*, "noL_0e =", noL_0e
PROVIDE noL_1e
PROVIDE noL_2e
endif

View File

@ -9,3 +9,14 @@ interface: ezfio
doc: Coefficients for the right wave function
type: double precision
size: (determinants.n_det,determinants.n_states)
[tc_gs_energy]
type: Threshold
doc: TC GS Energy
interface: ezfio
[tc_gs_var]
type: Threshold
doc: TC GS VAR
interface: ezfio

View File

@ -6,17 +6,38 @@ program print_tc_energy
implicit none
print *, 'Hello world'
read_wf = .True.
touch read_wf
my_grid_becke = .True.
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
read_wf = .True.
touch read_wf
call write_int(6, my_n_pt_r_grid, 'radial external grid over')
call write_int(6, my_n_pt_a_grid, 'angular external grid over')
if(tc_integ_type .eq. "numeric") then
my_extra_grid_becke = .True.
PROVIDE tc_grid2_a tc_grid2_r
my_n_pt_r_extra_grid = tc_grid2_r
my_n_pt_a_extra_grid = tc_grid2_a
touch my_extra_grid_becke my_n_pt_r_extra_grid my_n_pt_a_extra_grid
call write_int(6, my_n_pt_r_extra_grid, 'radial internal grid over')
call write_int(6, my_n_pt_a_extra_grid, 'angular internal grid over')
endif
call main()
end
! ---
subroutine main()
implicit none
PROVIDE j2e_type
PROVIDE j1e_type

View File

@ -6,7 +6,8 @@ program print_tc_var
implicit none
print *, 'Hello world'
print *, ' TC VAR is available only for HF REF WF'
print *, ' DO NOT FORGET TO RUN A CISD CALCULATION BEF'
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
@ -17,7 +18,7 @@ program print_tc_var
read_wf = .True.
touch read_wf
call write_tc_var()
call write_tc_gs_var_HF()
end

View File

@ -38,9 +38,9 @@ subroutine main()
call ezfio_has_cisd_energy(exists)
if(.not.exists) then
call ezfio_has_tc_scf_bitc_energy(exists)
call ezfio_has_tc_scf_tcscf_energy(exists)
if(exists) then
call ezfio_get_tc_scf_bitc_energy(e_ref)
call ezfio_get_tc_scf_tcscf_energy(e_ref)
endif
else
@ -59,7 +59,7 @@ subroutine main()
close(iunit)
end subroutine main
end
! --
@ -89,7 +89,7 @@ subroutine write_lr_spindeterminants()
call ezfio_set_spindeterminants_psi_left_coef_matrix_values(buffer)
deallocate(buffer)
end subroutine write_lr_spindeterminants
end
! ---

View File

@ -2,12 +2,67 @@
subroutine write_tc_energy()
implicit none
integer :: i, j, k
double precision :: hmono, htwoe, hthree, htot
double precision :: E_TC, O_TC
double precision :: E_1e, E_2e, E_3e
integer :: i, j, k
double precision :: hmono, htwoe, hthree, htot
double precision :: E_TC, O_TC
double precision :: E_1e, E_2e, E_3e
double precision, allocatable :: E_TC_tmp(:), E_1e_tmp(:), E_2e_tmp(:), E_3e_tmp(:)
do k = 1, n_states
! GS
! ---
allocate(E_TC_tmp(N_det), E_1e_tmp(N_det), E_2e_tmp(N_det), E_3e_tmp(N_det))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE(i, j, hmono, htwoe, hthree, htot) &
!$OMP SHARED(N_det, psi_det, N_int, psi_l_coef_bi_ortho, psi_r_coef_bi_ortho, &
!$OMP E_TC_tmp, E_1e_tmp, E_2e_tmp, E_3e_tmp)
!$OMP DO
do i = 1, N_det
E_TC_tmp(i) = 0.d0
E_1e_tmp(i) = 0.d0
E_2e_tmp(i) = 0.d0
E_3e_tmp(i) = 0.d0
do j = 1, N_det
call htilde_mu_mat_bi_ortho_slow(psi_det(1,1,i), psi_det(1,1,j), N_int, hmono, htwoe, hthree, htot)
E_TC_tmp(i) = E_TC_tmp(i) + psi_l_coef_bi_ortho(i,1) * psi_r_coef_bi_ortho(j,1) * htot
E_1e_tmp(i) = E_1e_tmp(i) + psi_l_coef_bi_ortho(i,1) * psi_r_coef_bi_ortho(j,1) * hmono
E_2e_tmp(i) = E_2e_tmp(i) + psi_l_coef_bi_ortho(i,1) * psi_r_coef_bi_ortho(j,1) * htwoe
E_3e_tmp(i) = E_3e_tmp(i) + psi_l_coef_bi_ortho(i,1) * psi_r_coef_bi_ortho(j,1) * hthree
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
E_1e = 0.d0
E_2e = 0.d0
E_3e = 0.d0
E_TC = 0.d0
O_TC = 0.d0
do i = 1, N_det
E_1e = E_1e + E_1e_tmp(i)
E_2e = E_2e + E_2e_tmp(i)
E_3e = E_3e + E_3e_tmp(i)
E_TC = E_TC + E_TC_tmp(i)
O_TC = O_TC + psi_l_coef_bi_ortho(i,1) * psi_r_coef_bi_ortho(i,1)
enddo
print *, ' state :', 1
print *, " E_TC = ", E_TC / O_TC
print *, " E_1e = ", E_1e / O_TC
print *, " E_2e = ", E_2e / O_TC
print *, " E_3e = ", E_3e / O_TC
print *, " O_TC = ", O_TC
call ezfio_set_tc_bi_ortho_tc_gs_energy(E_TC/O_TC)
! ---
! ES
! ---
do k = 2, n_states
E_TC = 0.d0
E_1e = 0.d0
@ -37,6 +92,8 @@ subroutine write_tc_energy()
enddo
deallocate(E_TC_tmp, E_1e_tmp, E_2e_tmp, E_3e_tmp)
end
! ---
@ -66,3 +123,25 @@ end
! ---
subroutine write_tc_gs_var_HF()
implicit none
integer :: i, j, k
double precision :: hmono, htwoe, hthree, htot
double precision :: SIGMA_TC
SIGMA_TC = 0.d0
do j = 2, N_det
call htilde_mu_mat_bi_ortho_slow(psi_det(1,1,j), psi_det(1,1,1), N_int, hmono, htwoe, hthree, htot)
SIGMA_TC = SIGMA_TC + htot * htot
enddo
print *, " SIGMA_TC = ", SIGMA_TC
call ezfio_set_tc_bi_ortho_tc_gs_var(SIGMA_TC)
end
! ---

View File

@ -24,44 +24,12 @@ program test_tc_fock
!call routine_2
! call routine_3()
! call test_3e
call routine_tot
end
! ---
subroutine test_3e
implicit none
double precision :: integral_aaa,integral_aab,integral_abb,integral_bbb,accu
double precision :: hmono, htwoe, hthree, htot
call htilde_mu_mat_bi_ortho_slow(ref_bitmask, ref_bitmask, N_int, hmono, htwoe, hthree, htot)
print*,'hmono = ',hmono
print*,'htwoe = ',htwoe
print*,'hthree= ',hthree
print*,'htot = ',htot
print*,''
print*,''
print*,'TC_one= ',tc_hf_one_e_energy
print*,'TC_two= ',TC_HF_two_e_energy
print*,'TC_3e = ',diag_three_elem_hf
print*,'TC_tot= ',TC_HF_energy
print*,''
print*,''
call give_aaa_contrib(integral_aaa)
print*,'integral_aaa = ',integral_aaa
call give_aab_contrib(integral_aab)
print*,'integral_aab = ',integral_aab
call give_abb_contrib(integral_abb)
print*,'integral_abb = ',integral_abb
call give_bbb_contrib(integral_bbb)
print*,'integral_bbb = ',integral_bbb
accu = integral_aaa + integral_aab + integral_abb + integral_bbb
print*,'accu = ',accu
print*,'delta = ',hthree - accu
end
subroutine routine_3()
use bitmasks ! you need to include the bitmasks_module.f90 features
@ -86,7 +54,6 @@ subroutine routine_3()
do i = 1, elec_num_tab(s1)
do a = elec_num_tab(s1)+1, mo_num ! virtual
det_i = ref_bitmask
call do_single_excitation(det_i, i, a, s1, i_ok)
if(i_ok == -1) then

View File

@ -100,30 +100,12 @@ doc: If |true|, the states are re-ordered to match the input states
default: False
interface: ezfio,provider,ocaml
[bi_ortho]
type: logical
doc: If |true|, the MO basis is assumed to be bi-orthonormal
interface: ezfio,provider,ocaml
default: True
[symmetric_fock_tc]
type: logical
doc: If |true|, using F+F^t as Fock TC
interface: ezfio,provider,ocaml
default: False
[thresh_tcscf]
type: Threshold
doc: Threshold on the convergence of the Hartree Fock energy.
interface: ezfio,provider,ocaml
default: 1.e-8
[n_it_tcscf_max]
type: Strictly_positive_int
doc: Maximum number of SCF iterations
interface: ezfio,provider,ocaml
default: 50
[selection_tc]
type: integer
doc: if +1: only positive is selected, -1: only negative is selected, :0 both positive and negative
@ -160,30 +142,6 @@ doc: If |true|, maximize the overlap between orthogonalized left- and right eige
interface: ezfio,provider,ocaml
default: False
[max_dim_diis_tcscf]
type: integer
doc: Maximum size of the DIIS extrapolation procedure
interface: ezfio,provider,ocaml
default: 15
[level_shift_tcscf]
type: Positive_float
doc: Energy shift on the virtual MOs to improve TCSCF convergence
interface: ezfio,provider,ocaml
default: 0.
[tcscf_algorithm]
type: character*(32)
doc: Type of TCSCF algorithm used. Possible choices are [Simple | DIIS]
interface: ezfio,provider,ocaml
default: DIIS
[im_thresh_tcscf]
type: Threshold
doc: Thresholds on the Imag part of energy
interface: ezfio,provider,ocaml
default: 1.e-7
[test_cycle_tc]
type: logical
doc: If |true|, the integrals of the three-body jastrow are computed with cycles
@ -304,3 +262,9 @@ doc: If |true|, more calc but less mem
interface: ezfio,provider,ocaml
default: False
[im_thresh_tc]
type: Threshold
doc: Thresholds on the Imag part of TC energy
interface: ezfio,provider,ocaml
default: 1.e-7

View File

@ -1,7 +0,0 @@
program tc_keywords
implicit none
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
print *, 'Hello world'
end

View File

@ -1,6 +1,6 @@
[bitc_energy]
[tcscf_energy]
type: Threshold
doc: Energy bi-tc HF
doc: TC-SCF ENERGY
interface: ezfio
[converged_tcscf]
@ -9,3 +9,33 @@ doc: If |true|, tc-scf has converged
interface: ezfio,provider,ocaml
default: False
[max_dim_diis_tcscf]
type: integer
doc: Maximum size of the DIIS extrapolation procedure
interface: ezfio,provider,ocaml
default: 15
[level_shift_tcscf]
type: Positive_float
doc: Energy shift on the virtual MOs to improve TCSCF convergence
interface: ezfio,provider,ocaml
default: 0.
[thresh_tcscf]
type: Threshold
doc: Threshold on the convergence of the Hartree Fock energy.
interface: ezfio,provider,ocaml
default: 1.e-8
[n_it_tcscf_max]
type: Strictly_positive_int
doc: Maximum number of SCF iterations
interface: ezfio,provider,ocaml
default: 50
[tc_Brillouin_Right]
type: logical
doc: If |true|, impose only right-Brillouin condition
interface: ezfio,provider,ocaml
default: False

View File

@ -1,75 +0,0 @@
! ---
program combine_lr_tcscf
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
implicit none
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
bi_ortho = .True.
touch bi_ortho
call comb_orbitals()
end
! ---
subroutine comb_orbitals()
implicit none
integer :: i, m, n, nn, mm
double precision :: accu_d, accu_nd
double precision, allocatable :: R(:,:), L(:,:), Rnew(:,:), tmp(:,:), S(:,:)
n = ao_num
m = mo_num
nn = elec_alpha_num
mm = m - nn
allocate(L(n,m), R(n,m), Rnew(n,m), S(m,m))
L = mo_l_coef
R = mo_r_coef
call check_weighted_biorthog(n, m, ao_overlap, L, R, accu_d, accu_nd, S, .true.)
allocate(tmp(n,nn))
do i = 1, nn
tmp(1:n,i) = R(1:n,i)
enddo
call impose_weighted_orthog_svd(n, nn, ao_overlap, tmp)
do i = 1, nn
Rnew(1:n,i) = tmp(1:n,i)
enddo
deallocate(tmp)
allocate(tmp(n,mm))
do i = 1, mm
tmp(1:n,i) = L(1:n,i+nn)
enddo
call impose_weighted_orthog_svd(n, mm, ao_overlap, tmp)
do i = 1, mm
Rnew(1:n,i+nn) = tmp(1:n,i)
enddo
deallocate(tmp)
call check_weighted_biorthog(n, m, ao_overlap, Rnew, Rnew, accu_d, accu_nd, S, .true.)
mo_r_coef = Rnew
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)
deallocate(L, R, Rnew, S)
end subroutine comb_orbitals
! ---

View File

@ -1,96 +0,0 @@
! ---
BEGIN_PROVIDER [ double precision, fock_vartc_eigvec_mo, (mo_num, mo_num)]
implicit none
integer :: i, j
integer :: liwork, lwork, n, info
integer, allocatable :: iwork(:)
double precision, allocatable :: work(:), F(:,:), F_save(:,:)
double precision, allocatable :: diag(:)
PROVIDE mo_r_coef
PROVIDE Fock_matrix_vartc_mo_tot
allocate( F(mo_num,mo_num), F_save(mo_num,mo_num) )
allocate (diag(mo_num) )
do j = 1, mo_num
do i = 1, mo_num
F(i,j) = Fock_matrix_vartc_mo_tot(i,j)
enddo
enddo
! Insert level shift here
do i = elec_beta_num+1, elec_alpha_num
F(i,i) += 0.5d0 * level_shift_tcscf
enddo
do i = elec_alpha_num+1, mo_num
F(i,i) += level_shift_tcscf
enddo
n = mo_num
lwork = 1+6*n + 2*n*n
liwork = 3 + 5*n
allocate(work(lwork))
allocate(iwork(liwork) )
lwork = -1
liwork = -1
F_save = F
call dsyevd('V', 'U', mo_num, F, size(F, 1), diag, work, lwork, iwork, liwork, info)
if (info /= 0) then
print *, irp_here//' DSYEVD failed : ', info
stop 1
endif
lwork = int(work(1))
liwork = iwork(1)
deallocate(iwork)
deallocate(work)
allocate(work(lwork))
allocate(iwork(liwork) )
call dsyevd('V', 'U', mo_num, F, size(F, 1), diag, work, lwork, iwork, liwork, info)
deallocate(iwork)
if (info /= 0) then
F = F_save
call dsyev('V', 'L', mo_num, F, size(F, 1), diag, work, lwork, info)
if (info /= 0) then
print *, irp_here//' DSYEV failed : ', info
stop 1
endif
endif
do i = 1, mo_num
do j = 1, mo_num
fock_vartc_eigvec_mo(j,i) = F(j,i)
enddo
enddo
deallocate(work, F, F_save, diag)
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, fock_vartc_eigvec_ao, (ao_num, mo_num)]
implicit none
PROVIDE mo_r_coef
call dgemm( 'N', 'N', ao_num, mo_num, mo_num, 1.d0 &
, mo_r_coef, size(mo_r_coef, 1), fock_vartc_eigvec_mo, size(fock_vartc_eigvec_mo, 1) &
, 0.d0, fock_vartc_eigvec_ao, size(fock_vartc_eigvec_ao, 1))
END_PROVIDER
! ---

View File

@ -91,28 +91,14 @@ BEGIN_PROVIDER [double precision, FQS_SQF_ao, (ao_num, ao_num)]
double precision, allocatable :: tmp(:,:)
double precision, allocatable :: F(:,:)
!print *, ' Providing FQS_SQF_ao ...'
!call wall_time(t0)
PROVIDE Fock_matrix_tc_ao_tot
allocate(F(ao_num,ao_num))
if(var_tc) then
do i = 1, ao_num
do j = 1, ao_num
F(j,i) = Fock_matrix_vartc_ao_tot(j,i)
enddo
do i = 1, ao_num
do j = 1, ao_num
F(j,i) = Fock_matrix_tc_ao_tot(j,i)
enddo
else
PROVIDE Fock_matrix_tc_ao_tot
do i = 1, ao_num
do j = 1, ao_num
F(j,i) = Fock_matrix_tc_ao_tot(j,i)
enddo
enddo
endif
enddo
allocate(tmp(ao_num,ao_num))
@ -140,9 +126,6 @@ BEGIN_PROVIDER [double precision, FQS_SQF_ao, (ao_num, ao_num)]
deallocate(tmp)
deallocate(F)
!call wall_time(t1)
!print *, ' Wall time for FQS_SQF_ao =', t1-t0
END_PROVIDER
! ---
@ -152,61 +135,13 @@ BEGIN_PROVIDER [double precision, FQS_SQF_mo, (mo_num, mo_num)]
implicit none
double precision :: t0, t1
!print*, ' Providing FQS_SQF_mo ...'
!call wall_time(t0)
PROVIDE mo_r_coef mo_l_coef
PROVIDE FQS_SQF_ao
call ao_to_mo_bi_ortho( FQS_SQF_ao, size(FQS_SQF_ao, 1) &
, FQS_SQF_mo, size(FQS_SQF_mo, 1) )
!call wall_time(t1)
!print*, ' Wall time for FQS_SQF_mo =', t1-t0
END_PROVIDER
! ---
! BEGIN_PROVIDER [ double precision, eigenval_Fock_tc_ao, (ao_num) ]
!&BEGIN_PROVIDER [ double precision, eigenvec_Fock_tc_ao, (ao_num,ao_num) ]
!
! BEGIN_DOC
! !
! ! Eigenvalues and eigenvectors of the Fock matrix over the ao basis
! !
! ! F' = X.T x F x X where X = ao_overlap^(-1/2)
! !
! ! F' x Cr' = Cr' x E ==> F Cr = Cr x E with Cr = X x Cr'
! ! F'.T x Cl' = Cl' x E ==> F.T Cl = Cl x E with Cl = X x Cl'
! !
! END_DOC
!
! implicit none
! double precision, allocatable :: tmp1(:,:), tmp2(:,:)
!
! ! ---
! ! Fock matrix in orthogonal basis: F' = X.T x F x X
!
! allocate(tmp1(ao_num,ao_num))
! call dgemm( 'N', 'N', ao_num, ao_num, ao_num, 1.d0 &
! , Fock_matrix_tc_ao_tot, size(Fock_matrix_tc_ao_tot, 1), S_half_inv, size(S_half_inv, 1) &
! , 0.d0, tmp1, size(tmp1, 1) )
!
! allocate(tmp2(ao_num,ao_num))
! call dgemm( 'T', 'N', ao_num, ao_num, ao_num, 1.d0 &
! , S_half_inv, size(S_half_inv, 1), tmp1, size(tmp1, 1) &
! , 0.d0, tmp2, size(tmp2, 1) )
!
! ! ---
!
! ! Diagonalize F' to obtain eigenvectors in orthogonal basis C' and eigenvalues
! ! TODO
!
! ! Back-transform eigenvectors: C =X.C'
!
!END_PROVIDER
! ---
~

View File

@ -1,299 +0,0 @@
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_cs, (mo_num, mo_num)]
implicit none
integer :: a, b, i, j, ipoint
double precision :: ti, tf
double precision :: loc_1, loc_2, loc_3
double precision, allocatable :: Okappa(:), Jkappa(:,:)
double precision, allocatable :: tmp_omp_d1(:), tmp_omp_d2(:,:)
double precision, allocatable :: tmp_1(:,:), tmp_2(:,:,:,:), tmp_22(:,:,:)
double precision, allocatable :: tmp_3(:,:,:), tmp_4(:,:,:)
PROVIDE mo_l_coef mo_r_coef
!print *, ' PROVIDING fock_3e_uhf_mo_cs ...'
!call wall_time(ti)
! ---
allocate(Jkappa(n_points_final_grid,3), Okappa(n_points_final_grid))
Jkappa = 0.d0
Okappa = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, i, tmp_omp_d1, tmp_omp_d2) &
!$OMP SHARED (n_points_final_grid, elec_beta_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, Okappa, Jkappa)
allocate(tmp_omp_d2(n_points_final_grid,3), tmp_omp_d1(n_points_final_grid))
tmp_omp_d2 = 0.d0
tmp_omp_d1 = 0.d0
!$OMP DO
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_omp_d2(ipoint,1) += int2_grad1_u12_bimo_t(ipoint,1,i,i)
tmp_omp_d2(ipoint,2) += int2_grad1_u12_bimo_t(ipoint,2,i,i)
tmp_omp_d2(ipoint,3) += int2_grad1_u12_bimo_t(ipoint,3,i,i)
tmp_omp_d1(ipoint) += mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do ipoint = 1, n_points_final_grid
Jkappa(ipoint,1) += tmp_omp_d2(ipoint,1)
Jkappa(ipoint,2) += tmp_omp_d2(ipoint,2)
Jkappa(ipoint,3) += tmp_omp_d2(ipoint,3)
Okappa(ipoint) += tmp_omp_d1(ipoint)
enddo
!$OMP END CRITICAL
deallocate(tmp_omp_d2, tmp_omp_d1)
!$OMP END PARALLEL
! ---
allocate(tmp_1(n_points_final_grid,4))
do ipoint = 1, n_points_final_grid
loc_1 = 2.d0 * Okappa(ipoint)
tmp_1(ipoint,1) = loc_1 * Jkappa(ipoint,1)
tmp_1(ipoint,2) = loc_1 * Jkappa(ipoint,2)
tmp_1(ipoint,3) = loc_1 * Jkappa(ipoint,3)
tmp_1(ipoint,4) = Okappa(ipoint)
enddo
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, i, j, loc_1, tmp_omp_d2) &
!$OMP SHARED (n_points_final_grid, elec_beta_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp_1)
allocate(tmp_omp_d2(n_points_final_grid,3))
tmp_omp_d2 = 0.d0
!$OMP DO COLLAPSE(2)
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
loc_1 = mos_l_in_r_array_transp(ipoint,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_omp_d2(ipoint,1) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,1,i,j)
tmp_omp_d2(ipoint,2) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,2,i,j)
tmp_omp_d2(ipoint,3) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,3,i,j)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do ipoint = 1, n_points_final_grid
tmp_1(ipoint,1) += tmp_omp_d2(ipoint,1)
tmp_1(ipoint,2) += tmp_omp_d2(ipoint,2)
tmp_1(ipoint,3) += tmp_omp_d2(ipoint,3)
enddo
!$OMP END CRITICAL
deallocate(tmp_omp_d2)
!$OMP END PARALLEL
! ---
if(tc_save_mem) then
allocate(tmp_22(n_points_final_grid,4,mo_num))
do a = 1, mo_num
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, a, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, final_weight_at_r_vector, &
!$OMP tmp_22)
!$OMP DO
do b = 1, mo_num
do ipoint = 1, n_points_final_grid
tmp_22(ipoint,1,b) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,1,b,a)
tmp_22(ipoint,2,b) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,2,b,a)
tmp_22(ipoint,3,b) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,3,b,a)
enddo
tmp_22(:,4,b) = 0.d0
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_22(ipoint,4,b) -= final_weight_at_r_vector(ipoint) * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,a) )
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
call dgemv( 'T', 4*n_points_final_grid, mo_num, -2.d0 &
, tmp_22(1,1,1), size(tmp_22, 1) * size(tmp_22, 2) &
, tmp_1(1,1), 1 &
, 0.d0, fock_3e_uhf_mo_cs(1,a), 1)
enddo
deallocate(tmp_22)
else
allocate(tmp_2(n_points_final_grid,4,mo_num,mo_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, a, b, i) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, final_weight_at_r_vector, &
!$OMP tmp_2)
!$OMP DO COLLAPSE(2)
do a = 1, mo_num
do b = 1, mo_num
do ipoint = 1, n_points_final_grid
tmp_2(ipoint,1,b,a) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,1,b,a)
tmp_2(ipoint,2,b,a) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,2,b,a)
tmp_2(ipoint,3,b,a) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,3,b,a)
enddo
tmp_2(:,4,b,a) = 0.d0
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_2(ipoint,4,b,a) -= final_weight_at_r_vector(ipoint) * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,a) )
enddo
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
call dgemv( 'T', 4*n_points_final_grid, mo_num*mo_num, -2.d0 &
, tmp_2(1,1,1,1), size(tmp_2, 1) * size(tmp_2, 2) &
, tmp_1(1,1), 1 &
, 0.d0, fock_3e_uhf_mo_cs(1,1), 1)
deallocate(tmp_2)
endif
deallocate(tmp_1)
! ---
allocate(tmp_3(n_points_final_grid,5,mo_num), tmp_4(n_points_final_grid,5,mo_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, loc_1, loc_2) &
!$OMP SHARED (n_points_final_grid, mo_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP final_weight_at_r_vector, Jkappa, tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
tmp_3(:,:,b) = 0.d0
tmp_4(:,:,b) = 0.d0
do ipoint = 1, n_points_final_grid
tmp_3(ipoint,1,b) = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,b)
tmp_4(ipoint,1,b) = -2.d0 * mos_r_in_r_array_transp(ipoint,b) * ( Jkappa(ipoint,1) * Jkappa(ipoint,1) &
+ Jkappa(ipoint,2) * Jkappa(ipoint,2) &
+ Jkappa(ipoint,3) * Jkappa(ipoint,3) )
tmp_4(ipoint,5,b) = mos_r_in_r_array_transp(ipoint,b)
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i, loc_1, loc_2) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP Jkappa, tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,i)
loc_2 = mos_r_in_r_array_transp(ipoint,i)
tmp_3(ipoint,2,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,1,b,i)
tmp_3(ipoint,3,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,2,b,i)
tmp_3(ipoint,4,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,3,b,i)
tmp_3(ipoint,5,b) += 2.d0 * loc_1 * ( Jkappa(ipoint,1) * int2_grad1_u12_bimo_t(ipoint,1,b,i) &
+ Jkappa(ipoint,2) * int2_grad1_u12_bimo_t(ipoint,2,b,i) &
+ Jkappa(ipoint,3) * int2_grad1_u12_bimo_t(ipoint,3,b,i) )
tmp_4(ipoint,2,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,1,i,b)
tmp_4(ipoint,3,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,2,i,b)
tmp_4(ipoint,4,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,3,i,b)
tmp_4(ipoint,1,b) += 2.d0 * loc_2 * ( Jkappa(ipoint,1) * int2_grad1_u12_bimo_t(ipoint,1,i,b) &
+ Jkappa(ipoint,2) * int2_grad1_u12_bimo_t(ipoint,2,i,b) &
+ Jkappa(ipoint,3) * int2_grad1_u12_bimo_t(ipoint,3,i,b) )
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i, j, loc_1, loc_2, loc_3) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,j)
loc_2 = mos_r_in_r_array_transp(ipoint,b)
loc_3 = mos_r_in_r_array_transp(ipoint,i)
tmp_3(ipoint,5,b) -= loc_1 * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,j) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,j) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,j) )
tmp_4(ipoint,1,b) += ( loc_2 * ( int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i) ) &
- loc_3 * ( int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,b) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,b) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,b) ) )
enddo
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
! ---
call dgemm( 'T', 'N', mo_num, mo_num, 5*n_points_final_grid, 1.d0 &
, tmp_3(1,1,1), 5*n_points_final_grid &
, tmp_4(1,1,1), 5*n_points_final_grid &
, 1.d0, fock_3e_uhf_mo_cs(1,1), mo_num)
deallocate(tmp_3, tmp_4)
deallocate(Jkappa, Okappa)
! ---
!call wall_time(tf)
!print *, ' total Wall time for fock_3e_uhf_mo_cs =', (tf - ti) / 60.d0
END_PROVIDER
! ---

View File

@ -1,536 +0,0 @@
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_a_os, (mo_num, mo_num)]
&BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_b_os, (mo_num, mo_num)]
BEGIN_DOC
!
! Open Shell part of the Fock matrix from three-electron terms
!
! WARNING :: non hermitian if bi-ortho MOS used
!
END_DOC
implicit none
integer :: a, b, i, j, ipoint
double precision :: loc_1, loc_2, loc_3, loc_4
double precision :: ti, tf
double precision, allocatable :: Okappa(:), Jkappa(:,:), Obarkappa(:), Jbarkappa(:,:)
double precision, allocatable :: tmp_omp_d1(:), tmp_omp_d2(:,:)
double precision, allocatable :: tmp_1(:,:), tmp_2(:,:,:,:)
double precision, allocatable :: tmp_3(:,:,:), tmp_4(:,:,:)
PROVIDE mo_l_coef mo_r_coef
!print *, ' Providing fock_3e_uhf_mo_a_os and fock_3e_uhf_mo_b_os ...'
!call wall_time(ti)
! ---
allocate(Jkappa(n_points_final_grid,3), Okappa(n_points_final_grid))
allocate(Jbarkappa(n_points_final_grid,3), Obarkappa(n_points_final_grid))
Jkappa = 0.d0
Okappa = 0.d0
Jbarkappa = 0.d0
Obarkappa = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, i, tmp_omp_d1, tmp_omp_d2) &
!$OMP SHARED (n_points_final_grid, elec_beta_num, elec_alpha_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, Okappa, Jkappa, Obarkappa, Jbarkappa)
allocate(tmp_omp_d2(n_points_final_grid,3), tmp_omp_d1(n_points_final_grid))
tmp_omp_d2 = 0.d0
tmp_omp_d1 = 0.d0
!$OMP DO
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_omp_d2(ipoint,1) += int2_grad1_u12_bimo_t(ipoint,1,i,i)
tmp_omp_d2(ipoint,2) += int2_grad1_u12_bimo_t(ipoint,2,i,i)
tmp_omp_d2(ipoint,3) += int2_grad1_u12_bimo_t(ipoint,3,i,i)
tmp_omp_d1(ipoint) += mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do ipoint = 1, n_points_final_grid
Jkappa(ipoint,1) += tmp_omp_d2(ipoint,1)
Jkappa(ipoint,2) += tmp_omp_d2(ipoint,2)
Jkappa(ipoint,3) += tmp_omp_d2(ipoint,3)
Okappa(ipoint) += tmp_omp_d1(ipoint)
enddo
!$OMP END CRITICAL
tmp_omp_d2 = 0.d0
tmp_omp_d1 = 0.d0
!$OMP DO
do i = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
tmp_omp_d2(ipoint,1) += int2_grad1_u12_bimo_t(ipoint,1,i,i)
tmp_omp_d2(ipoint,2) += int2_grad1_u12_bimo_t(ipoint,2,i,i)
tmp_omp_d2(ipoint,3) += int2_grad1_u12_bimo_t(ipoint,3,i,i)
tmp_omp_d1(ipoint) += mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do ipoint = 1, n_points_final_grid
Jbarkappa(ipoint,1) += tmp_omp_d2(ipoint,1)
Jbarkappa(ipoint,2) += tmp_omp_d2(ipoint,2)
Jbarkappa(ipoint,3) += tmp_omp_d2(ipoint,3)
Obarkappa(ipoint) += tmp_omp_d1(ipoint)
enddo
!$OMP END CRITICAL
deallocate(tmp_omp_d2, tmp_omp_d1)
!$OMP END PARALLEL
! ---
allocate(tmp_1(n_points_final_grid,4))
do ipoint = 1, n_points_final_grid
loc_1 = -2.d0 * Okappa (ipoint)
loc_2 = -2.d0 * Obarkappa(ipoint)
loc_3 = Obarkappa(ipoint)
tmp_1(ipoint,1) = (loc_1 - loc_3) * Jbarkappa(ipoint,1) + loc_2 * Jkappa(ipoint,1)
tmp_1(ipoint,2) = (loc_1 - loc_3) * Jbarkappa(ipoint,2) + loc_2 * Jkappa(ipoint,2)
tmp_1(ipoint,3) = (loc_1 - loc_3) * Jbarkappa(ipoint,3) + loc_2 * Jkappa(ipoint,3)
tmp_1(ipoint,4) = Obarkappa(ipoint)
enddo
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, i, j, loc_1, loc_2, tmp_omp_d2) &
!$OMP SHARED (n_points_final_grid, elec_beta_num, elec_alpha_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp_1)
allocate(tmp_omp_d2(n_points_final_grid,3))
tmp_omp_d2 = 0.d0
!$OMP DO COLLAPSE(2)
do i = 1, elec_beta_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_1 = mos_l_in_r_array_transp(ipoint,j) * mos_r_in_r_array_transp(ipoint,i)
loc_2 = mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_omp_d2(ipoint,1) += loc_1 * int2_grad1_u12_bimo_t(ipoint,1,i,j) + loc_2 * int2_grad1_u12_bimo_t(ipoint,1,j,i)
tmp_omp_d2(ipoint,2) += loc_1 * int2_grad1_u12_bimo_t(ipoint,2,i,j) + loc_2 * int2_grad1_u12_bimo_t(ipoint,2,j,i)
tmp_omp_d2(ipoint,3) += loc_1 * int2_grad1_u12_bimo_t(ipoint,3,i,j) + loc_2 * int2_grad1_u12_bimo_t(ipoint,3,j,i)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do ipoint = 1, n_points_final_grid
tmp_1(ipoint,1) += tmp_omp_d2(ipoint,1)
tmp_1(ipoint,2) += tmp_omp_d2(ipoint,2)
tmp_1(ipoint,3) += tmp_omp_d2(ipoint,3)
enddo
!$OMP END CRITICAL
tmp_omp_d2 = 0.d0
!$OMP DO COLLAPSE(2)
do i = elec_beta_num+1, elec_alpha_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_1 = mos_l_in_r_array_transp(ipoint,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_omp_d2(ipoint,1) += loc_1 * int2_grad1_u12_bimo_t(ipoint,1,i,j)
tmp_omp_d2(ipoint,2) += loc_1 * int2_grad1_u12_bimo_t(ipoint,2,i,j)
tmp_omp_d2(ipoint,3) += loc_1 * int2_grad1_u12_bimo_t(ipoint,3,i,j)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do ipoint = 1, n_points_final_grid
tmp_1(ipoint,1) += tmp_omp_d2(ipoint,1)
tmp_1(ipoint,2) += tmp_omp_d2(ipoint,2)
tmp_1(ipoint,3) += tmp_omp_d2(ipoint,3)
enddo
!$OMP END CRITICAL
deallocate(tmp_omp_d2)
!$OMP END PARALLEL
! ---
allocate(tmp_2(n_points_final_grid,4,mo_num,mo_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, a, b) &
!$OMP SHARED (n_points_final_grid, mo_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, final_weight_at_r_vector, &
!$OMP tmp_2)
!$OMP DO COLLAPSE(2)
do a = 1, mo_num
do b = 1, mo_num
do ipoint = 1, n_points_final_grid
tmp_2(ipoint,1,b,a) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,1,b,a)
tmp_2(ipoint,2,b,a) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,2,b,a)
tmp_2(ipoint,3,b,a) = final_weight_at_r_vector(ipoint) * int2_grad1_u12_bimo_t(ipoint,3,b,a)
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, a, b, i) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, elec_alpha_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP tmp_2)
!$OMP DO COLLAPSE(2)
do a = 1, mo_num
do b = 1, mo_num
tmp_2(:,4,b,a) = 0.d0
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_2(ipoint,4,b,a) += final_weight_at_r_vector(ipoint) * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,a) )
enddo
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
! ---
call dgemv( 'T', 4*n_points_final_grid, mo_num*mo_num, 1.d0 &
, tmp_2(1,1,1,1), size(tmp_2, 1) * size(tmp_2, 2) &
, tmp_1(1,1), 1 &
, 0.d0, fock_3e_uhf_mo_b_os(1,1), 1)
deallocate(tmp_1, tmp_2)
! ---
allocate(tmp_3(n_points_final_grid,2,mo_num), tmp_4(n_points_final_grid,2,mo_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, loc_1, loc_2) &
!$OMP SHARED (n_points_final_grid, mo_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP final_weight_at_r_vector, Jkappa, Jbarkappa, tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
tmp_3(:,:,b) = 0.d0
tmp_4(:,:,b) = 0.d0
do ipoint = 1, n_points_final_grid
tmp_3(ipoint,1,b) = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,b)
loc_1 = -2.0d0 * mos_r_in_r_array_transp(ipoint,b)
tmp_4(ipoint,1,b) = loc_1 * ( Jbarkappa(ipoint,1) * (Jkappa(ipoint,1) + 0.25d0 * Jbarkappa(ipoint,1)) &
+ Jbarkappa(ipoint,2) * (Jkappa(ipoint,2) + 0.25d0 * Jbarkappa(ipoint,2)) &
+ Jbarkappa(ipoint,3) * (Jkappa(ipoint,3) + 0.25d0 * Jbarkappa(ipoint,3)) )
tmp_4(ipoint,2,b) = mos_r_in_r_array_transp(ipoint,b)
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i, loc_1, loc_2, loc_3, loc_4) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, elec_alpha_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP Jkappa, Jbarkappa, tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,i)
loc_2 = mos_r_in_r_array_transp(ipoint,i)
tmp_3(ipoint,2,b) += loc_1 * ( Jbarkappa(ipoint,1) * int2_grad1_u12_bimo_t(ipoint,1,b,i) &
+ Jbarkappa(ipoint,2) * int2_grad1_u12_bimo_t(ipoint,2,b,i) &
+ Jbarkappa(ipoint,3) * int2_grad1_u12_bimo_t(ipoint,3,b,i) )
tmp_4(ipoint,1,b) += loc_2 * ( Jbarkappa(ipoint,1) * int2_grad1_u12_bimo_t(ipoint,1,i,b) &
+ Jbarkappa(ipoint,2) * int2_grad1_u12_bimo_t(ipoint,2,i,b) &
+ Jbarkappa(ipoint,3) * int2_grad1_u12_bimo_t(ipoint,3,i,b) )
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i, j, loc_1, loc_2, loc_3) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, elec_alpha_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
do i = 1, elec_beta_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_2 = mos_r_in_r_array_transp(ipoint,b)
tmp_4(ipoint,1,b) += loc_2 * ( int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i) )
enddo
enddo
enddo
do i = elec_beta_num+1, elec_alpha_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_2 = 0.5d0 * mos_r_in_r_array_transp(ipoint,b)
tmp_4(ipoint,1,b) += loc_2 * ( int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i) )
enddo
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
! ---
call dgemm( 'T', 'N', mo_num, mo_num, 2*n_points_final_grid, 1.d0 &
, tmp_3(1,1,1), 2*n_points_final_grid &
, tmp_4(1,1,1), 2*n_points_final_grid &
, 1.d0, fock_3e_uhf_mo_b_os(1,1), mo_num)
deallocate(tmp_3, tmp_4)
! ---
fock_3e_uhf_mo_a_os = fock_3e_uhf_mo_b_os
allocate(tmp_1(n_points_final_grid,1))
do ipoint = 1, n_points_final_grid
tmp_1(ipoint,1) = Obarkappa(ipoint) + 2.d0 * Okappa(ipoint)
enddo
allocate(tmp_2(n_points_final_grid,1,mo_num,mo_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, a, b, i) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, elec_alpha_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP tmp_2)
!$OMP DO COLLAPSE(2)
do a = 1, mo_num
do b = 1, mo_num
tmp_2(:,1,b,a) = 0.d0
do i = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
tmp_2(ipoint,1,b,a) += final_weight_at_r_vector(ipoint) * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,a) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,a) )
enddo
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
call dgemv( 'T', n_points_final_grid, mo_num*mo_num, 1.d0 &
, tmp_2(1,1,1,1), size(tmp_2, 1) * size(tmp_2, 2) &
, tmp_1(1,1), 1 &
, 1.d0, fock_3e_uhf_mo_a_os(1,1), 1)
deallocate(tmp_1, tmp_2)
! ---
allocate(tmp_3(n_points_final_grid,8,mo_num), tmp_4(n_points_final_grid,8,mo_num))
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b) &
!$OMP SHARED (n_points_final_grid, mo_num, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP final_weight_at_r_vector, Jkappa, Jbarkappa, tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
tmp_3(:,:,b) = 0.d0
tmp_4(:,:,b) = 0.d0
do ipoint = 1, n_points_final_grid
tmp_3(ipoint,1,b) = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,b)
tmp_4(ipoint,8,b) = mos_r_in_r_array_transp(ipoint,b)
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i, loc_1, loc_2, loc_3, loc_4) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, elec_alpha_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP Jkappa, Jbarkappa, tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,i)
loc_2 = mos_r_in_r_array_transp(ipoint,i)
tmp_3(ipoint,2,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,1,b,i)
tmp_3(ipoint,3,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,2,b,i)
tmp_3(ipoint,4,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,3,b,i)
tmp_4(ipoint,5,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,1,i,b)
tmp_4(ipoint,6,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,2,i,b)
tmp_4(ipoint,7,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,3,i,b)
enddo
enddo
do i = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,i)
loc_3 = 2.d0 * loc_1
loc_2 = mos_r_in_r_array_transp(ipoint,i)
loc_4 = 2.d0 * loc_2
tmp_3(ipoint,5,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,1,b,i)
tmp_3(ipoint,6,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,2,b,i)
tmp_3(ipoint,7,b) -= loc_1 * int2_grad1_u12_bimo_t(ipoint,3,b,i)
tmp_3(ipoint,8,b) += loc_3 * ( (Jkappa(ipoint,1) + 0.5d0 * Jbarkappa(ipoint,1)) * int2_grad1_u12_bimo_t(ipoint,1,b,i) &
+ (Jkappa(ipoint,2) + 0.5d0 * Jbarkappa(ipoint,2)) * int2_grad1_u12_bimo_t(ipoint,2,b,i) &
+ (Jkappa(ipoint,3) + 0.5d0 * Jbarkappa(ipoint,3)) * int2_grad1_u12_bimo_t(ipoint,3,b,i) )
tmp_4(ipoint,1,b) += loc_4 * ( (Jkappa(ipoint,1) + 0.5d0 * Jbarkappa(ipoint,1)) * int2_grad1_u12_bimo_t(ipoint,1,i,b) &
+ (Jkappa(ipoint,2) + 0.5d0 * Jbarkappa(ipoint,2)) * int2_grad1_u12_bimo_t(ipoint,2,i,b) &
+ (Jkappa(ipoint,3) + 0.5d0 * Jbarkappa(ipoint,3)) * int2_grad1_u12_bimo_t(ipoint,3,i,b) )
tmp_4(ipoint,2,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,1,i,b)
tmp_4(ipoint,3,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,2,i,b)
tmp_4(ipoint,4,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,3,i,b)
tmp_4(ipoint,5,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,1,i,b)
tmp_4(ipoint,6,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,2,i,b)
tmp_4(ipoint,7,b) += loc_2 * int2_grad1_u12_bimo_t(ipoint,3,i,b)
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
!$OMP PARALLEL &
!$OMP DEFAULT (NONE) &
!$OMP PRIVATE (ipoint, b, i, j, loc_1, loc_2, loc_3) &
!$OMP SHARED (n_points_final_grid, mo_num, elec_beta_num, elec_alpha_num, &
!$OMP final_weight_at_r_vector, int2_grad1_u12_bimo_t, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP tmp_3, tmp_4)
!$OMP DO
do b = 1, mo_num
do i = 1, elec_beta_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,j)
loc_2 = mos_r_in_r_array_transp(ipoint,b)
loc_3 = mos_r_in_r_array_transp(ipoint,i)
tmp_3(ipoint,8,b) -= loc_1 * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,j) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,j) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,j) )
tmp_4(ipoint,1,b) -= loc_3 * ( int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,b) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,b) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,b) )
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,i)
loc_3 = mos_r_in_r_array_transp(ipoint,j)
tmp_3(ipoint,8,b) -= loc_1 * ( int2_grad1_u12_bimo_t(ipoint,1,b,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i) )
tmp_4(ipoint,1,b) -= loc_3 * ( int2_grad1_u12_bimo_t(ipoint,1,j,i) * int2_grad1_u12_bimo_t(ipoint,1,i,b) &
+ int2_grad1_u12_bimo_t(ipoint,2,j,i) * int2_grad1_u12_bimo_t(ipoint,2,i,b) &
+ int2_grad1_u12_bimo_t(ipoint,3,j,i) * int2_grad1_u12_bimo_t(ipoint,3,i,b) )
enddo
enddo
enddo
do i = elec_beta_num+1, elec_alpha_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
loc_1 = final_weight_at_r_vector(ipoint) * mos_l_in_r_array_transp(ipoint,j)
loc_2 = 0.5d0 * mos_r_in_r_array_transp(ipoint,b)
loc_3 = mos_r_in_r_array_transp(ipoint,i)
tmp_3(ipoint,8,b) -= loc_1 * ( int2_grad1_u12_bimo_t(ipoint,1,b,i) * int2_grad1_u12_bimo_t(ipoint,1,i,j) &
+ int2_grad1_u12_bimo_t(ipoint,2,b,i) * int2_grad1_u12_bimo_t(ipoint,2,i,j) &
+ int2_grad1_u12_bimo_t(ipoint,3,b,i) * int2_grad1_u12_bimo_t(ipoint,3,i,j) )
tmp_4(ipoint,1,b) -= loc_3 * ( int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,b) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,b) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,b) )
enddo
enddo
enddo
enddo
!$OMP END DO
!$OMP END PARALLEL
! ---
call dgemm( 'T', 'N', mo_num, mo_num, 8*n_points_final_grid, 1.d0 &
, tmp_3(1,1,1), 8*n_points_final_grid &
, tmp_4(1,1,1), 8*n_points_final_grid &
, 1.d0, fock_3e_uhf_mo_a_os(1,1), mo_num)
deallocate(tmp_3, tmp_4)
deallocate(Jkappa, Okappa)
!call wall_time(tf)
!print *, ' Wall time for fock_3e_uhf_mo_a_os and fock_3e_uhf_mo_b_os =', tf - ti
END_PROVIDER
! ---

View File

@ -1,77 +0,0 @@
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_a, (mo_num, mo_num)]
BEGIN_DOC
!
! Fock matrix alpha from three-electron terms
!
! WARNING :: non hermitian if bi-ortho MOS used
!
END_DOC
implicit none
double precision :: ti, tf
PROVIDE mo_l_coef mo_r_coef
!print *, ' Providing fock_3e_uhf_mo_a ...'
!call wall_time(ti)
! CLOSED-SHELL PART
PROVIDE fock_3e_uhf_mo_cs
fock_3e_uhf_mo_a = fock_3e_uhf_mo_cs
if(elec_alpha_num .ne. elec_beta_num) then
! OPEN-SHELL PART
PROVIDE fock_3e_uhf_mo_a_os
fock_3e_uhf_mo_a += fock_3e_uhf_mo_a_os
endif
!call wall_time(tf)
!print *, ' Wall time for fock_3e_uhf_mo_a (min) =', (tf - ti)/60.d0
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_b, (mo_num, mo_num)]
BEGIN_DOC
!
! Fock matrix beta from three-electron terms
!
! WARNING :: non hermitian if bi-ortho MOS used
!
END_DOC
implicit none
double precision :: ti, tf
PROVIDE mo_l_coef mo_r_coef
!print *, ' Providing and fock_3e_uhf_mo_b ...'
!call wall_time(ti)
! CLOSED-SHELL PART
PROVIDE fock_3e_uhf_mo_cs
fock_3e_uhf_mo_b = fock_3e_uhf_mo_cs
if(elec_alpha_num .ne. elec_beta_num) then
! OPEN-SHELL PART
PROVIDE fock_3e_uhf_mo_b_os
fock_3e_uhf_mo_b += fock_3e_uhf_mo_b_os
endif
!call wall_time(tf)
!print *, ' Wall time for fock_3e_uhf_mo_b =', tf - ti
END_PROVIDER
! ---

View File

@ -1,490 +0,0 @@
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_cs_old, (mo_num, mo_num)]
implicit none
integer :: a, b, i, j
double precision :: I_bij_aij, I_bij_ija, I_bij_jai, I_bij_aji, I_bij_iaj, I_bij_jia
double precision :: ti, tf
double precision, allocatable :: tmp(:,:)
PROVIDE mo_l_coef mo_r_coef
call give_integrals_3_body_bi_ort(1, 1, 1, 1, 1, 1, I_bij_aij)
!print *, ' PROVIDING fock_3e_uhf_mo_cs_old ...'
!call wall_time(ti)
fock_3e_uhf_mo_cs_old = 0.d0
!$OMP PARALLEL DEFAULT (NONE) &
!$OMP PRIVATE (a, b, i, j, I_bij_aij, I_bij_ija, I_bij_jai, I_bij_aji, I_bij_iaj, I_bij_jia, tmp) &
!$OMP SHARED (mo_num, elec_beta_num, fock_3e_uhf_mo_cs_old)
allocate(tmp(mo_num,mo_num))
tmp = 0.d0
!$OMP DO
do a = 1, mo_num
do b = 1, mo_num
do j = 1, elec_beta_num
do i = 1, elec_beta_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( 4.d0 * I_bij_aij &
+ I_bij_ija &
+ I_bij_jai &
- 2.d0 * I_bij_aji &
- 2.d0 * I_bij_iaj &
- 2.d0 * I_bij_jia )
enddo
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do a = 1, mo_num
do b = 1, mo_num
fock_3e_uhf_mo_cs_old(b,a) += tmp(b,a)
enddo
enddo
!$OMP END CRITICAL
deallocate(tmp)
!$OMP END PARALLEL
!call wall_time(tf)
!print *, ' total Wall time for fock_3e_uhf_mo_cs_old =', tf - ti
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_a_old, (mo_num, mo_num)]
BEGIN_DOC
!
! ALPHA part of the Fock matrix from three-electron terms
!
! WARNING :: non hermitian if bi-ortho MOS used
!
END_DOC
implicit none
integer :: a, b, i, j, o
double precision :: I_bij_aij, I_bij_ija, I_bij_jai, I_bij_aji, I_bij_iaj, I_bij_jia
double precision :: ti, tf
double precision, allocatable :: tmp(:,:)
PROVIDE mo_l_coef mo_r_coef
PROVIDE fock_3e_uhf_mo_cs
!print *, ' Providing fock_3e_uhf_mo_a_old ...'
!call wall_time(ti)
o = elec_beta_num + 1
call give_integrals_3_body_bi_ort(1, 1, 1, 1, 1, 1, I_bij_aij)
PROVIDE fock_3e_uhf_mo_cs_old
fock_3e_uhf_mo_a_old = fock_3e_uhf_mo_cs_old
!$OMP PARALLEL DEFAULT (NONE) &
!$OMP PRIVATE (a, b, i, j, I_bij_aij, I_bij_ija, I_bij_jai, I_bij_aji, I_bij_iaj, I_bij_jia, tmp) &
!$OMP SHARED (mo_num, o, elec_alpha_num, elec_beta_num, fock_3e_uhf_mo_a_old)
allocate(tmp(mo_num,mo_num))
tmp = 0.d0
!$OMP DO
do a = 1, mo_num
do b = 1, mo_num
! ---
do j = o, elec_alpha_num
do i = 1, elec_beta_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( 2.d0 * I_bij_aij &
+ I_bij_ija &
+ I_bij_jai &
- I_bij_aji &
- I_bij_iaj &
- 2.d0 * I_bij_jia )
enddo
enddo
! ---
do j = 1, elec_beta_num
do i = o, elec_alpha_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( 2.d0 * I_bij_aij &
+ I_bij_ija &
+ I_bij_jai &
- I_bij_aji &
- 2.d0 * I_bij_iaj &
- I_bij_jia )
enddo
enddo
! ---
do j = o, elec_alpha_num
do i = o, elec_alpha_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( I_bij_aij &
+ I_bij_ija &
+ I_bij_jai &
- I_bij_aji &
- I_bij_iaj &
- I_bij_jia )
enddo
enddo
! ---
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do a = 1, mo_num
do b = 1, mo_num
fock_3e_uhf_mo_a_old(b,a) += tmp(b,a)
enddo
enddo
!$OMP END CRITICAL
deallocate(tmp)
!$OMP END PARALLEL
!call wall_time(tf)
!print *, ' Wall time for fock_3e_uhf_mo_a_old =', tf - ti
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_mo_b_old, (mo_num, mo_num)]
BEGIN_DOC
!
! BETA part of the Fock matrix from three-electron terms
!
! WARNING :: non hermitian if bi-ortho MOS used
!
END_DOC
implicit none
integer :: a, b, i, j, o
double precision :: I_bij_aij, I_bij_ija, I_bij_jai, I_bij_aji, I_bij_iaj, I_bij_jia
double precision :: ti, tf
double precision, allocatable :: tmp(:,:)
PROVIDE mo_l_coef mo_r_coef
!print *, ' PROVIDING fock_3e_uhf_mo_b_old ...'
!call wall_time(ti)
o = elec_beta_num + 1
call give_integrals_3_body_bi_ort(1, 1, 1, 1, 1, 1, I_bij_aij)
PROVIDE fock_3e_uhf_mo_cs_old
fock_3e_uhf_mo_b_old = fock_3e_uhf_mo_cs_old
!$OMP PARALLEL DEFAULT (NONE) &
!$OMP PRIVATE (a, b, i, j, I_bij_aij, I_bij_ija, I_bij_jai, I_bij_aji, I_bij_iaj, I_bij_jia, tmp) &
!$OMP SHARED (mo_num, o, elec_alpha_num, elec_beta_num, fock_3e_uhf_mo_b_old)
allocate(tmp(mo_num,mo_num))
tmp = 0.d0
!$OMP DO
do a = 1, mo_num
do b = 1, mo_num
! ---
do j = o, elec_alpha_num
do i = 1, elec_beta_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( 2.d0 * I_bij_aij &
- I_bij_aji &
- I_bij_iaj )
enddo
enddo
! ---
do j = 1, elec_beta_num
do i = o, elec_alpha_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( 2.d0 * I_bij_aij &
- I_bij_aji &
- I_bij_jia )
enddo
enddo
! ---
do j = o, elec_alpha_num
do i = o, elec_alpha_num
call give_integrals_3_body_bi_ort(b, i, j, a, i, j, I_bij_aij)
call give_integrals_3_body_bi_ort(b, i, j, i, j, a, I_bij_ija)
call give_integrals_3_body_bi_ort(b, i, j, j, a, i, I_bij_jai)
call give_integrals_3_body_bi_ort(b, i, j, a, j, i, I_bij_aji)
call give_integrals_3_body_bi_ort(b, i, j, i, a, j, I_bij_iaj)
call give_integrals_3_body_bi_ort(b, i, j, j, i, a, I_bij_jia)
tmp(b,a) -= 0.5d0 * ( I_bij_aij &
- I_bij_aji )
enddo
enddo
! ---
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do a = 1, mo_num
do b = 1, mo_num
fock_3e_uhf_mo_b_old(b,a) += tmp(b,a)
enddo
enddo
!$OMP END CRITICAL
deallocate(tmp)
!$OMP END PARALLEL
!call wall_time(tf)
!print *, ' total Wall time for fock_3e_uhf_mo_b_old =', tf - ti
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_ao_a, (ao_num, ao_num)]
BEGIN_DOC
!
! Equations (B6) and (B7)
!
! g <--> gamma
! d <--> delta
! e <--> eta
! k <--> kappa
!
END_DOC
implicit none
integer :: g, d, e, k, mu, nu
double precision :: dm_ge_a, dm_ge_b, dm_ge
double precision :: dm_dk_a, dm_dk_b, dm_dk
double precision :: i_mugd_nuek, i_mugd_eknu, i_mugd_knue, i_mugd_nuke, i_mugd_enuk, i_mugd_kenu
double precision :: ti, tf
double precision, allocatable :: f_tmp(:,:)
!print *, ' PROVIDING fock_3e_uhf_ao_a ...'
!call wall_time(ti)
fock_3e_uhf_ao_a = 0.d0
!$OMP PARALLEL DEFAULT (NONE) &
!$OMP PRIVATE (g, e, d, k, mu, nu, dm_ge_a, dm_ge_b, dm_ge, dm_dk_a, dm_dk_b, dm_dk, f_tmp, &
!$OMP i_mugd_nuek, i_mugd_eknu, i_mugd_knue, i_mugd_nuke, i_mugd_enuk, i_mugd_kenu) &
!$OMP SHARED (ao_num, TCSCF_bi_ort_dm_ao_alpha, TCSCF_bi_ort_dm_ao_beta, fock_3e_uhf_ao_a)
allocate(f_tmp(ao_num,ao_num))
f_tmp = 0.d0
!$OMP DO
do g = 1, ao_num
do e = 1, ao_num
dm_ge_a = TCSCF_bi_ort_dm_ao_alpha(g,e)
dm_ge_b = TCSCF_bi_ort_dm_ao_beta (g,e)
dm_ge = dm_ge_a + dm_ge_b
do d = 1, ao_num
do k = 1, ao_num
dm_dk_a = TCSCF_bi_ort_dm_ao_alpha(d,k)
dm_dk_b = TCSCF_bi_ort_dm_ao_beta (d,k)
dm_dk = dm_dk_a + dm_dk_b
do mu = 1, ao_num
do nu = 1, ao_num
call give_integrals_3_body_bi_ort_ao(mu, g, d, nu, e, k, i_mugd_nuek)
call give_integrals_3_body_bi_ort_ao(mu, g, d, e, k, nu, i_mugd_eknu)
call give_integrals_3_body_bi_ort_ao(mu, g, d, k, nu, e, i_mugd_knue)
call give_integrals_3_body_bi_ort_ao(mu, g, d, nu, k, e, i_mugd_nuke)
call give_integrals_3_body_bi_ort_ao(mu, g, d, e, nu, k, i_mugd_enuk)
call give_integrals_3_body_bi_ort_ao(mu, g, d, k, e, nu, i_mugd_kenu)
f_tmp(mu,nu) -= 0.5d0 * ( dm_ge * dm_dk * i_mugd_nuek &
+ dm_ge_a * dm_dk_a * i_mugd_eknu &
+ dm_ge_a * dm_dk_a * i_mugd_knue &
- dm_ge_a * dm_dk * i_mugd_enuk &
- dm_ge * dm_dk_a * i_mugd_kenu &
- dm_ge_a * dm_dk_a * i_mugd_nuke &
- dm_ge_b * dm_dk_b * i_mugd_nuke )
enddo
enddo
enddo
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do mu = 1, ao_num
do nu = 1, ao_num
fock_3e_uhf_ao_a(mu,nu) += f_tmp(mu,nu)
enddo
enddo
!$OMP END CRITICAL
deallocate(f_tmp)
!$OMP END PARALLEL
!call wall_time(tf)
!print *, ' total Wall time for fock_3e_uhf_ao_a =', tf - ti
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, fock_3e_uhf_ao_b, (ao_num, ao_num)]
BEGIN_DOC
!
! Equations (B6) and (B7)
!
! g <--> gamma
! d <--> delta
! e <--> eta
! k <--> kappa
!
END_DOC
implicit none
integer :: g, d, e, k, mu, nu
double precision :: dm_ge_a, dm_ge_b, dm_ge
double precision :: dm_dk_a, dm_dk_b, dm_dk
double precision :: i_mugd_nuek, i_mugd_eknu, i_mugd_knue, i_mugd_nuke, i_mugd_enuk, i_mugd_kenu
double precision :: ti, tf
double precision, allocatable :: f_tmp(:,:)
!print *, ' PROVIDING fock_3e_uhf_ao_b ...'
!call wall_time(ti)
fock_3e_uhf_ao_b = 0.d0
!$OMP PARALLEL DEFAULT (NONE) &
!$OMP PRIVATE (g, e, d, k, mu, nu, dm_ge_a, dm_ge_b, dm_ge, dm_dk_a, dm_dk_b, dm_dk, f_tmp, &
!$OMP i_mugd_nuek, i_mugd_eknu, i_mugd_knue, i_mugd_nuke, i_mugd_enuk, i_mugd_kenu) &
!$OMP SHARED (ao_num, TCSCF_bi_ort_dm_ao_alpha, TCSCF_bi_ort_dm_ao_beta, fock_3e_uhf_ao_b)
allocate(f_tmp(ao_num,ao_num))
f_tmp = 0.d0
!$OMP DO
do g = 1, ao_num
do e = 1, ao_num
dm_ge_a = TCSCF_bi_ort_dm_ao_alpha(g,e)
dm_ge_b = TCSCF_bi_ort_dm_ao_beta (g,e)
dm_ge = dm_ge_a + dm_ge_b
do d = 1, ao_num
do k = 1, ao_num
dm_dk_a = TCSCF_bi_ort_dm_ao_alpha(d,k)
dm_dk_b = TCSCF_bi_ort_dm_ao_beta (d,k)
dm_dk = dm_dk_a + dm_dk_b
do mu = 1, ao_num
do nu = 1, ao_num
call give_integrals_3_body_bi_ort_ao(mu, g, d, nu, e, k, i_mugd_nuek)
call give_integrals_3_body_bi_ort_ao(mu, g, d, e, k, nu, i_mugd_eknu)
call give_integrals_3_body_bi_ort_ao(mu, g, d, k, nu, e, i_mugd_knue)
call give_integrals_3_body_bi_ort_ao(mu, g, d, nu, k, e, i_mugd_nuke)
call give_integrals_3_body_bi_ort_ao(mu, g, d, e, nu, k, i_mugd_enuk)
call give_integrals_3_body_bi_ort_ao(mu, g, d, k, e, nu, i_mugd_kenu)
f_tmp(mu,nu) -= 0.5d0 * ( dm_ge * dm_dk * i_mugd_nuek &
+ dm_ge_b * dm_dk_b * i_mugd_eknu &
+ dm_ge_b * dm_dk_b * i_mugd_knue &
- dm_ge_b * dm_dk * i_mugd_enuk &
- dm_ge * dm_dk_b * i_mugd_kenu &
- dm_ge_b * dm_dk_b * i_mugd_nuke &
- dm_ge_a * dm_dk_a * i_mugd_nuke )
enddo
enddo
enddo
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do mu = 1, ao_num
do nu = 1, ao_num
fock_3e_uhf_ao_b(mu,nu) += f_tmp(mu,nu)
enddo
enddo
!$OMP END CRITICAL
deallocate(f_tmp)
!$OMP END PARALLEL
!call wall_time(tf)
!print *, ' total Wall time for fock_3e_uhf_ao_b =', tf - ti
END_PROVIDER
! ---

View File

@ -1,107 +0,0 @@
! ---
BEGIN_PROVIDER [ double precision, good_hermit_tc_fock_mat, (mo_num, mo_num)]
BEGIN_DOC
! good_hermit_tc_fock_mat = Hermitian Upper triangular Fock matrix
!
! The converged eigenvectors of such matrix yield to orthonormal vectors satisfying the left Brillouin theorem
END_DOC
implicit none
integer :: i, j
good_hermit_tc_fock_mat = Fock_matrix_tc_mo_tot
do j = 1, mo_num
do i = 1, j-1
good_hermit_tc_fock_mat(i,j) = Fock_matrix_tc_mo_tot(j,i)
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, hermit_average_tc_fock_mat, (mo_num, mo_num)]
BEGIN_DOC
! hermit_average_tc_fock_mat = (F + F^\dagger)/2
END_DOC
implicit none
integer :: i, j
hermit_average_tc_fock_mat = Fock_matrix_tc_mo_tot
do j = 1, mo_num
do i = 1, mo_num
hermit_average_tc_fock_mat(i,j) = 0.5d0 * (Fock_matrix_tc_mo_tot(j,i) + Fock_matrix_tc_mo_tot(i,j))
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, grad_hermit]
implicit none
BEGIN_DOC
! square of gradient of the energy
END_DOC
if(symmetric_fock_tc)then
grad_hermit = grad_hermit_average_tc_fock_mat
else
grad_hermit = grad_good_hermit_tc_fock_mat
endif
END_PROVIDER
BEGIN_PROVIDER [ double precision, grad_good_hermit_tc_fock_mat]
implicit none
BEGIN_DOC
! grad_good_hermit_tc_fock_mat = norm of gradients of the upper triangular TC fock
END_DOC
integer :: i, j
grad_good_hermit_tc_fock_mat = 0.d0
do i = 1, elec_alpha_num
do j = elec_alpha_num+1, mo_num
grad_good_hermit_tc_fock_mat += dabs(good_hermit_tc_fock_mat(i,j))
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, grad_hermit_average_tc_fock_mat]
implicit none
BEGIN_DOC
! grad_hermit_average_tc_fock_mat = norm of gradients of the upper triangular TC fock
END_DOC
integer :: i, j
grad_hermit_average_tc_fock_mat = 0.d0
do i = 1, elec_alpha_num
do j = elec_alpha_num+1, mo_num
grad_hermit_average_tc_fock_mat += dabs(hermit_average_tc_fock_mat(i,j))
enddo
enddo
END_PROVIDER
! ---
subroutine save_good_hermit_tc_eigvectors()
implicit none
integer :: sign
character*(64) :: label
logical :: output
sign = 1
label = "Canonical"
output = .False.
if(symmetric_fock_tc)then
call mo_as_eigvectors_of_mo_matrix(hermit_average_tc_fock_mat, mo_num, mo_num, label, sign, output)
else
call mo_as_eigvectors_of_mo_matrix(good_hermit_tc_fock_mat, mo_num, mo_num, label, sign, output)
endif
end subroutine save_good_hermit_tc_eigvectors
! ---

File diff suppressed because it is too large Load Diff

View File

@ -1,4 +1,6 @@
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_tc_mo_tot, (mo_num,mo_num) ]
&BEGIN_PROVIDER [ double precision, Fock_matrix_tc_diag_mo_tot, (mo_num)]
@ -23,9 +25,6 @@
integer :: i, j, n
double precision :: t0, t1
!print*, ' Providing Fock_matrix_tc_mo_tot ...'
!call wall_time(t0)
if(elec_alpha_num == elec_beta_num) then
PROVIDE Fock_matrix_tc_mo_alpha
@ -133,7 +132,7 @@
enddo
endif
if(no_oa_or_av_opt)then
if(no_oa_or_av_opt) then
do i = 1, n_act_orb
iorb = list_act(i)
do j = 1, n_inact_orb
@ -154,12 +153,25 @@
enddo
endif
if(.not.bi_ortho .and. three_body_h_tc)then
Fock_matrix_tc_mo_tot += fock_3_mat
endif
if(tc_Brillouin_Right) then
!call wall_time(t1)
!print*, ' Wall time for Fock_matrix_tc_mo_tot =', t1-t0
double precision, allocatable :: tmp(:,:)
allocate(tmp(mo_num,mo_num))
tmp = Fock_matrix_tc_mo_tot
do j = 1, mo_num
do i = 1, j-1
tmp(i,j) = Fock_matrix_tc_mo_tot(j,i)
enddo
enddo
Fock_matrix_tc_mo_tot = tmp
deallocate(tmp)
endif
END_PROVIDER
! ---

View File

@ -1,771 +0,0 @@
! ---
BEGIN_PROVIDER [ double precision, fock_3_mat, (mo_num, mo_num)]
implicit none
integer :: i,j
double precision :: contrib
fock_3_mat = 0.d0
if(.not.bi_ortho .and. three_body_h_tc) then
call give_fock_ia_three_e_total(1, 1, contrib)
!! !$OMP PARALLEL &
!! !$OMP DEFAULT (NONE) &
!! !$OMP PRIVATE (i,j,m,integral) &
!! !$OMP SHARED (mo_num,three_body_3_index)
!! !$OMP DO SCHEDULE (guided) COLLAPSE(3)
do i = 1, mo_num
do j = 1, mo_num
call give_fock_ia_three_e_total(j,i,contrib)
fock_3_mat(j,i) = -contrib
enddo
enddo
!else if(bi_ortho.and.three_body_h_tc) then
!! !$OMP END DO
!! !$OMP END PARALLEL
!! do i = 1, mo_num
!! do j = 1, i-1
!! mat_three(j,i) = mat_three(i,j)
!! enddo
!! enddo
endif
END_PROVIDER
subroutine give_fock_ia_three_e_total(i,a,contrib)
implicit none
BEGIN_DOC
! contrib is the TOTAL (same spins / opposite spins) contribution from the three body term to the Fock operator
!
END_DOC
integer, intent(in) :: i,a
double precision, intent(out) :: contrib
double precision :: int_1, int_2, int_3
double precision :: mos_i, mos_a, w_ia
double precision :: mos_ia, weight
integer :: mm, ipoint,k,l
int_1 = 0.d0
int_2 = 0.d0
int_3 = 0.d0
do mm = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
mos_i = mos_in_r_array_transp(ipoint,i)
mos_a = mos_in_r_array_transp(ipoint,a)
mos_ia = mos_a * mos_i
w_ia = x_W_ij_erf_rk(ipoint,mm,i,a)
int_1 += weight * fock_3_w_kk_sum(ipoint,mm) * (4.d0 * fock_3_rho_beta(ipoint) * w_ia &
+ 2.0d0 * mos_ia * fock_3_w_kk_sum(ipoint,mm) &
- 2.0d0 * fock_3_w_ki_mos_k(ipoint,mm,i) * mos_a &
- 2.0d0 * fock_3_w_ki_mos_k(ipoint,mm,a) * mos_i )
int_2 += weight * (-1.d0) * ( 2.0d0 * fock_3_w_kl_mo_k_mo_l(ipoint,mm) * w_ia &
+ 2.0d0 * fock_3_rho_beta(ipoint) * fock_3_w_ki_wk_a(ipoint,mm,i,a) &
+ 1.0d0 * mos_ia * fock_3_trace_w_tilde(ipoint,mm) )
int_3 += weight * 1.d0 * (fock_3_w_kl_wla_phi_k(ipoint,mm,i) * mos_a + fock_3_w_kl_wla_phi_k(ipoint,mm,a) * mos_i &
+fock_3_w_ki_mos_k(ipoint,mm,i) * fock_3_w_ki_mos_k(ipoint,mm,a) )
enddo
enddo
contrib = int_1 + int_2 + int_3
end
! ---
BEGIN_PROVIDER [double precision, diag_three_elem_hf]
implicit none
integer :: i, j, k, ipoint, mm
double precision :: contrib, weight, four_third, one_third, two_third, exchange_int_231
double precision :: integral_aaa, hthree, integral_aab, integral_abb, integral_bbb
double precision, allocatable :: tmp(:)
double precision, allocatable :: tmp_L(:,:), tmp_R(:,:)
double precision, allocatable :: tmp_M(:,:), tmp_S(:), tmp_O(:), tmp_J(:,:)
double precision, allocatable :: tmp_M_priv(:,:), tmp_S_priv(:), tmp_O_priv(:), tmp_J_priv(:,:)
PROVIDE mo_l_coef mo_r_coef
!print *, ' providing diag_three_elem_hf'
if(.not. three_body_h_tc) then
if(noL_standard) then
PROVIDE noL_0e
diag_three_elem_hf = noL_0e
else
diag_three_elem_hf = 0.d0
endif
else
if(.not. bi_ortho) then
! ---
one_third = 1.d0/3.d0
two_third = 2.d0/3.d0
four_third = 4.d0/3.d0
diag_three_elem_hf = 0.d0
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do k = 1, elec_beta_num
call give_integrals_3_body(k, j, i, j, i, k, exchange_int_231)
diag_three_elem_hf += two_third * exchange_int_231
enddo
enddo
enddo
do mm = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
contrib = 3.d0 * fock_3_w_kk_sum(ipoint,mm) * fock_3_rho_beta(ipoint) * fock_3_w_kk_sum(ipoint,mm) &
- 2.d0 * fock_3_w_kl_mo_k_mo_l(ipoint,mm) * fock_3_w_kk_sum(ipoint,mm) &
- 1.d0 * fock_3_rho_beta(ipoint) * fock_3_w_kl_w_kl(ipoint,mm)
contrib *= four_third
contrib += -two_third * fock_3_rho_beta(ipoint) * fock_3_w_kl_w_kl(ipoint,mm) &
-four_third * fock_3_w_kk_sum(ipoint,mm) * fock_3_w_kl_mo_k_mo_l(ipoint,mm)
diag_three_elem_hf += weight * contrib
enddo
enddo
diag_three_elem_hf = - diag_three_elem_hf
! ---
else
! ------------
! SLOW VERSION
! ------------
!call give_aaa_contrib(integral_aaa)
!call give_aab_contrib(integral_aab)
!call give_abb_contrib(integral_abb)
!call give_bbb_contrib(integral_bbb)
!diag_three_elem_hf = integral_aaa + integral_aab + integral_abb + integral_bbb
! ------------
! ------------
PROVIDE int2_grad1_u12_bimo_t
PROVIDE mos_l_in_r_array_transp
PROVIDE mos_r_in_r_array_transp
if(elec_alpha_num .eq. elec_beta_num) then
allocate(tmp(elec_beta_num))
allocate(tmp_L(n_points_final_grid,3), tmp_R(n_points_final_grid,3))
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(j, i, ipoint, tmp_L, tmp_R) &
!$OMP SHARED(elec_beta_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp, final_weight_at_r_vector)
!$OMP DO
do j = 1, elec_beta_num
tmp_L = 0.d0
tmp_R = 0.d0
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_L(ipoint,1) = tmp_L(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,2) = tmp_L(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,3) = tmp_L(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_R(ipoint,1) = tmp_R(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,2) = tmp_R(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,3) = tmp_R(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,i,j) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
tmp(j) = 0.d0
do ipoint = 1, n_points_final_grid
tmp(j) = tmp(j) + final_weight_at_r_vector(ipoint) * (tmp_L(ipoint,1)*tmp_R(ipoint,1) + tmp_L(ipoint,2)*tmp_R(ipoint,2) + tmp_L(ipoint,3)*tmp_R(ipoint,3))
enddo
enddo ! j
!$OMP END DO
!$OMP END PARALLEL
diag_three_elem_hf = -2.d0 * sum(tmp)
deallocate(tmp)
deallocate(tmp_L, tmp_R)
! ---
allocate(tmp_O(n_points_final_grid), tmp_J(n_points_final_grid,3))
tmp_O = 0.d0
tmp_J = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(i, ipoint, tmp_O_priv, tmp_J_priv) &
!$OMP SHARED(elec_beta_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp_O, tmp_J)
allocate(tmp_O_priv(n_points_final_grid), tmp_J_priv(n_points_final_grid,3))
tmp_O_priv = 0.d0
tmp_J_priv = 0.d0
!$OMP DO
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_O_priv(ipoint) = tmp_O_priv(ipoint) + mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,i)
tmp_J_priv(ipoint,1) = tmp_J_priv(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,i,i)
tmp_J_priv(ipoint,2) = tmp_J_priv(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,i,i)
tmp_J_priv(ipoint,3) = tmp_J_priv(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,i,i)
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
tmp_O = tmp_O + tmp_O_priv
tmp_J = tmp_J + tmp_J_priv
!$OMP END CRITICAL
deallocate(tmp_O_priv, tmp_J_priv)
!$OMP END PARALLEL
allocate(tmp_M(n_points_final_grid,3), tmp_S(n_points_final_grid))
tmp_M = 0.d0
tmp_S = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(i, j, ipoint, tmp_M_priv, tmp_S_priv) &
!$OMP SHARED(elec_beta_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp_M, tmp_S)
allocate(tmp_M_priv(n_points_final_grid,3), tmp_S_priv(n_points_final_grid))
tmp_M_priv = 0.d0
tmp_S_priv = 0.d0
!$OMP DO COLLAPSE(2)
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_M_priv(ipoint,1) = tmp_M_priv(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,2) = tmp_M_priv(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,3) = tmp_M_priv(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_S_priv(ipoint) = tmp_S_priv(ipoint) + int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
tmp_M = tmp_M + tmp_M_priv
tmp_S = tmp_S + tmp_S_priv
!$OMP END CRITICAL
deallocate(tmp_M_priv, tmp_S_priv)
!$OMP END PARALLEL
allocate(tmp(n_points_final_grid))
do ipoint = 1, n_points_final_grid
tmp_S(ipoint) = 2.d0 * (tmp_J(ipoint,1)*tmp_J(ipoint,1) + tmp_J(ipoint,2)*tmp_J(ipoint,2) + tmp_J(ipoint,3)*tmp_J(ipoint,3)) - tmp_S(ipoint)
tmp(ipoint) = final_weight_at_r_vector(ipoint) * ( tmp_O(ipoint) * tmp_S(ipoint) &
- 2.d0 * ( tmp_J(ipoint,1) * tmp_M(ipoint,1) &
+ tmp_J(ipoint,2) * tmp_M(ipoint,2) &
+ tmp_J(ipoint,3) * tmp_M(ipoint,3)))
enddo
diag_three_elem_hf = diag_three_elem_hf -2.d0 * (sum(tmp))
deallocate(tmp)
else
allocate(tmp(elec_alpha_num))
allocate(tmp_L(n_points_final_grid,3), tmp_R(n_points_final_grid,3))
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(j, i, ipoint, tmp_L, tmp_R) &
!$OMP SHARED(elec_beta_num, elec_alpha_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp, final_weight_at_r_vector)
!$OMP DO
do j = 1, elec_beta_num
tmp_L = 0.d0
tmp_R = 0.d0
do i = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
tmp_L(ipoint,1) = tmp_L(ipoint,1) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,2) = tmp_L(ipoint,2) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,3) = tmp_L(ipoint,3) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_R(ipoint,1) = tmp_R(ipoint,1) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,2) = tmp_R(ipoint,2) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,3) = tmp_R(ipoint,3) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,i,j) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
tmp(j) = 0.d0
do ipoint = 1, n_points_final_grid
tmp(j) = tmp(j) + final_weight_at_r_vector(ipoint) * (tmp_L(ipoint,1)*tmp_R(ipoint,1) + tmp_L(ipoint,2)*tmp_R(ipoint,2) + tmp_L(ipoint,3)*tmp_R(ipoint,3))
enddo
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_L(ipoint,1) = tmp_L(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,2) = tmp_L(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,3) = tmp_L(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_R(ipoint,1) = tmp_R(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,2) = tmp_R(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,3) = tmp_R(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,i,j) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
do ipoint = 1, n_points_final_grid
tmp(j) = tmp(j) + final_weight_at_r_vector(ipoint) * (tmp_L(ipoint,1)*tmp_R(ipoint,1) + tmp_L(ipoint,2)*tmp_R(ipoint,2) + tmp_L(ipoint,3)*tmp_R(ipoint,3))
enddo
enddo ! j
!$OMP END DO
!$OMP END PARALLEL
! ---
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(j, i, ipoint, tmp_L, tmp_R) &
!$OMP SHARED(elec_beta_num, elec_alpha_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp, final_weight_at_r_vector)
!$OMP DO
do j = elec_beta_num+1, elec_alpha_num
tmp_L = 0.d0
tmp_R = 0.d0
do i = 1, elec_alpha_num
do ipoint = 1, n_points_final_grid
tmp_L(ipoint,1) = tmp_L(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,2) = tmp_L(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_L(ipoint,3) = tmp_L(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i)
tmp_R(ipoint,1) = tmp_R(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,2) = tmp_R(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,i,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_R(ipoint,3) = tmp_R(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,i,j) * mos_r_in_r_array_transp(ipoint,i)
enddo
enddo
tmp(j) = 0.d0
do ipoint = 1, n_points_final_grid
tmp(j) = tmp(j) + 0.5d0 * final_weight_at_r_vector(ipoint) * (tmp_L(ipoint,1)*tmp_R(ipoint,1) + tmp_L(ipoint,2)*tmp_R(ipoint,2) + tmp_L(ipoint,3)*tmp_R(ipoint,3))
enddo
enddo ! j
!$OMP END DO
!$OMP END PARALLEL
diag_three_elem_hf = -2.d0 * sum(tmp)
deallocate(tmp)
deallocate(tmp_L, tmp_R)
! ---
allocate(tmp_O(n_points_final_grid), tmp_J(n_points_final_grid,3))
tmp_O = 0.d0
tmp_J = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(i, ipoint, tmp_O_priv, tmp_J_priv) &
!$OMP SHARED(elec_beta_num, elec_alpha_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp_O, tmp_J)
allocate(tmp_O_priv(n_points_final_grid), tmp_J_priv(n_points_final_grid,3))
tmp_O_priv = 0.d0
tmp_J_priv = 0.d0
!$OMP DO
do i = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_O_priv(ipoint) = tmp_O_priv(ipoint) + mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,i)
tmp_J_priv(ipoint,1) = tmp_J_priv(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,i,i)
tmp_J_priv(ipoint,2) = tmp_J_priv(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,i,i)
tmp_J_priv(ipoint,3) = tmp_J_priv(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,i,i)
enddo
enddo
!$OMP END DO NOWAIT
!$OMP DO
do i = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
tmp_O_priv(ipoint) = tmp_O_priv(ipoint) + 0.5d0 * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,i)
tmp_J_priv(ipoint,1) = tmp_J_priv(ipoint,1) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,i,i)
tmp_J_priv(ipoint,2) = tmp_J_priv(ipoint,2) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,i,i)
tmp_J_priv(ipoint,3) = tmp_J_priv(ipoint,3) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,i,i)
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
tmp_O = tmp_O + tmp_O_priv
tmp_J = tmp_J + tmp_J_priv
!$OMP END CRITICAL
deallocate(tmp_O_priv, tmp_J_priv)
!$OMP END PARALLEL
! ---
allocate(tmp_M(n_points_final_grid,3), tmp_S(n_points_final_grid))
tmp_M = 0.d0
tmp_S = 0.d0
!$OMP PARALLEL &
!$OMP DEFAULT(NONE) &
!$OMP PRIVATE(i, j, ipoint, tmp_M_priv, tmp_S_priv) &
!$OMP SHARED(elec_beta_num, elec_alpha_num, n_points_final_grid, &
!$OMP mos_l_in_r_array_transp, mos_r_in_r_array_transp, &
!$OMP int2_grad1_u12_bimo_t, tmp_M, tmp_S)
allocate(tmp_M_priv(n_points_final_grid,3), tmp_S_priv(n_points_final_grid))
tmp_M_priv = 0.d0
tmp_S_priv = 0.d0
!$OMP DO COLLAPSE(2)
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_M_priv(ipoint,1) = tmp_M_priv(ipoint,1) + int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,2) = tmp_M_priv(ipoint,2) + int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,3) = tmp_M_priv(ipoint,3) + int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_S_priv(ipoint) = tmp_S_priv(ipoint) + int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP DO COLLAPSE(2)
do i = elec_beta_num+1, elec_alpha_num
do j = 1, elec_beta_num
do ipoint = 1, n_points_final_grid
tmp_M_priv(ipoint,1) = tmp_M_priv(ipoint,1) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,2) = tmp_M_priv(ipoint,2) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,3) = tmp_M_priv(ipoint,3) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,1) = tmp_M_priv(ipoint,1) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,i,j) * mos_l_in_r_array_transp(ipoint,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_M_priv(ipoint,2) = tmp_M_priv(ipoint,2) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,i,j) * mos_l_in_r_array_transp(ipoint,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_M_priv(ipoint,3) = tmp_M_priv(ipoint,3) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,i,j) * mos_l_in_r_array_transp(ipoint,j) * mos_r_in_r_array_transp(ipoint,i)
tmp_S_priv(ipoint) = tmp_S_priv(ipoint) + int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP DO COLLAPSE(2)
do i = elec_beta_num+1, elec_alpha_num
do j = elec_beta_num+1, elec_alpha_num
do ipoint = 1, n_points_final_grid
tmp_M_priv(ipoint,1) = tmp_M_priv(ipoint,1) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,2) = tmp_M_priv(ipoint,2) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_M_priv(ipoint,3) = tmp_M_priv(ipoint,3) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,j,i) * mos_l_in_r_array_transp(ipoint,i) * mos_r_in_r_array_transp(ipoint,j)
tmp_S_priv(ipoint) = tmp_S_priv(ipoint) + 0.5d0 * int2_grad1_u12_bimo_t(ipoint,1,i,j) * int2_grad1_u12_bimo_t(ipoint,1,j,i) &
+ 0.5d0 * int2_grad1_u12_bimo_t(ipoint,2,i,j) * int2_grad1_u12_bimo_t(ipoint,2,j,i) &
+ 0.5d0 * int2_grad1_u12_bimo_t(ipoint,3,i,j) * int2_grad1_u12_bimo_t(ipoint,3,j,i)
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
tmp_M = tmp_M + tmp_M_priv
tmp_S = tmp_S + tmp_S_priv
!$OMP END CRITICAL
deallocate(tmp_M_priv, tmp_S_priv)
!$OMP END PARALLEL
allocate(tmp(n_points_final_grid))
do ipoint = 1, n_points_final_grid
tmp_S(ipoint) = 2.d0 * (tmp_J(ipoint,1)*tmp_J(ipoint,1) + tmp_J(ipoint,2)*tmp_J(ipoint,2) + tmp_J(ipoint,3)*tmp_J(ipoint,3)) - tmp_S(ipoint)
tmp(ipoint) = final_weight_at_r_vector(ipoint) * ( tmp_O(ipoint) * tmp_S(ipoint) &
- 2.d0 * ( tmp_J(ipoint,1) * tmp_M(ipoint,1) &
+ tmp_J(ipoint,2) * tmp_M(ipoint,2) &
+ tmp_J(ipoint,3) * tmp_M(ipoint,3)))
enddo
diag_three_elem_hf = diag_three_elem_hf - 2.d0 * (sum(tmp))
deallocate(tmp)
endif
endif
endif
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, fock_3_mat_a_op_sh, (mo_num, mo_num)]
implicit none
integer :: h,p,i,j
double precision :: direct_int, exch_int, exchange_int_231, exchange_int_312
double precision :: exchange_int_23, exchange_int_12, exchange_int_13
fock_3_mat_a_op_sh = 0.d0
do h = 1, mo_num
do p = 1, mo_num
!F_a^{ab}(h,p)
do i = 1, elec_beta_num ! beta
do j = elec_beta_num+1, elec_alpha_num ! alpha
call give_integrals_3_body(h,j,i,p,j,i,direct_int) ! <hji|pji>
call give_integrals_3_body(h,j,i,j,p,i,exch_int)
fock_3_mat_a_op_sh(h,p) -= direct_int - exch_int
enddo
enddo
!F_a^{aa}(h,p)
do i = 1, elec_beta_num ! alpha
do j = elec_beta_num+1, elec_alpha_num ! alpha
call give_integrals_3_body(h,j,i,p,j,i,direct_int)
call give_integrals_3_body(h,j,i,i,p,j,exchange_int_231)
call give_integrals_3_body(h,j,i,j,i,p,exchange_int_312)
call give_integrals_3_body(h,j,i,p,i,j,exchange_int_23)
call give_integrals_3_body(h,j,i,i,j,p,exchange_int_12)
call give_integrals_3_body(h,j,i,j,p,i,exchange_int_13)
fock_3_mat_a_op_sh(h,p) -= ( direct_int + exchange_int_231 + exchange_int_312 &
- exchange_int_23 & ! i <-> j
- exchange_int_12 & ! p <-> j
- exchange_int_13 )! p <-> i
enddo
enddo
enddo
enddo
! symmetrized
! do p = 1, elec_beta_num
! do h = elec_alpha_num +1, mo_num
! fock_3_mat_a_op_sh(h,p) = fock_3_mat_a_op_sh(p,h)
! enddo
! enddo
! do h = elec_beta_num+1, elec_alpha_num
! do p = elec_alpha_num +1, mo_num
! !F_a^{bb}(h,p)
! do i = 1, elec_beta_num
! do j = i+1, elec_beta_num
! call give_integrals_3_body(h,j,i,p,j,i,direct_int)
! call give_integrals_3_body(h,j,i,p,i,j,exch_int)
! fock_3_mat_a_op_sh(h,p) -= direct_int - exch_int
! enddo
! enddo
! enddo
! enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_mat_b_op_sh, (mo_num, mo_num)]
implicit none
integer :: h,p,i,j
double precision :: direct_int, exch_int
fock_3_mat_b_op_sh = 0.d0
do h = 1, elec_beta_num
do p = elec_alpha_num +1, mo_num
!F_b^{aa}(h,p)
do i = 1, elec_beta_num
do j = elec_beta_num+1, elec_alpha_num
call give_integrals_3_body(h,j,i,p,j,i,direct_int)
call give_integrals_3_body(h,j,i,p,i,j,exch_int)
fock_3_mat_b_op_sh(h,p) += direct_int - exch_int
enddo
enddo
!F_b^{ab}(h,p)
do i = elec_beta_num+1, elec_beta_num
do j = 1, elec_beta_num
call give_integrals_3_body(h,j,i,p,j,i,direct_int)
call give_integrals_3_body(h,j,i,j,p,i,exch_int)
fock_3_mat_b_op_sh(h,p) += direct_int - exch_int
enddo
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_w_kk_sum, (n_points_final_grid,3)]
implicit none
integer :: mm, ipoint,k
double precision :: w_kk
fock_3_w_kk_sum = 0.d0
do k = 1, elec_beta_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
w_kk = x_W_ij_erf_rk(ipoint,mm,k,k)
fock_3_w_kk_sum(ipoint,mm) += w_kk
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_w_ki_mos_k, (n_points_final_grid,3,mo_num)]
implicit none
integer :: mm, ipoint,k,i
double precision :: w_ki, mo_k
fock_3_w_ki_mos_k = 0.d0
do i = 1, mo_num
do k = 1, elec_beta_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
w_ki = x_W_ij_erf_rk(ipoint,mm,k,i)
mo_k = mos_in_r_array(k,ipoint)
fock_3_w_ki_mos_k(ipoint,mm,i) += w_ki * mo_k
enddo
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_w_kl_w_kl, (n_points_final_grid,3)]
implicit none
integer :: k,j,ipoint,mm
double precision :: w_kj
fock_3_w_kl_w_kl = 0.d0
do j = 1, elec_beta_num
do k = 1, elec_beta_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
w_kj = x_W_ij_erf_rk(ipoint,mm,k,j)
fock_3_w_kl_w_kl(ipoint,mm) += w_kj * w_kj
enddo
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_rho_beta, (n_points_final_grid)]
implicit none
integer :: ipoint,k
fock_3_rho_beta = 0.d0
do ipoint = 1, n_points_final_grid
do k = 1, elec_beta_num
fock_3_rho_beta(ipoint) += mos_in_r_array(k,ipoint) * mos_in_r_array(k,ipoint)
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_w_kl_mo_k_mo_l, (n_points_final_grid,3)]
implicit none
integer :: ipoint,k,l,mm
double precision :: mos_k, mos_l, w_kl
fock_3_w_kl_mo_k_mo_l = 0.d0
do k = 1, elec_beta_num
do l = 1, elec_beta_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
mos_k = mos_in_r_array_transp(ipoint,k)
mos_l = mos_in_r_array_transp(ipoint,l)
w_kl = x_W_ij_erf_rk(ipoint,mm,l,k)
fock_3_w_kl_mo_k_mo_l(ipoint,mm) += w_kl * mos_k * mos_l
enddo
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_w_ki_wk_a, (n_points_final_grid,3,mo_num, mo_num)]
implicit none
integer :: ipoint,i,a,k,mm
double precision :: w_ki,w_ka
fock_3_w_ki_wk_a = 0.d0
do i = 1, mo_num
do a = 1, mo_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
do k = 1, elec_beta_num
w_ki = x_W_ij_erf_rk(ipoint,mm,k,i)
w_ka = x_W_ij_erf_rk(ipoint,mm,k,a)
fock_3_w_ki_wk_a(ipoint,mm,a,i) += w_ki * w_ka
enddo
enddo
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_trace_w_tilde, (n_points_final_grid,3)]
implicit none
integer :: ipoint,k,mm
fock_3_trace_w_tilde = 0.d0
do k = 1, elec_beta_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
fock_3_trace_w_tilde(ipoint,mm) += fock_3_w_ki_wk_a(ipoint,mm,k,k)
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, fock_3_w_kl_wla_phi_k, (n_points_final_grid,3,mo_num)]
implicit none
integer :: ipoint,a,k,mm,l
double precision :: w_kl,w_la, mo_k
fock_3_w_kl_wla_phi_k = 0.d0
do a = 1, mo_num
do k = 1, elec_beta_num
do l = 1, elec_beta_num
do mm = 1, 3
do ipoint = 1, n_points_final_grid
w_kl = x_W_ij_erf_rk(ipoint,mm,l,k)
w_la = x_W_ij_erf_rk(ipoint,mm,l,a)
mo_k = mos_in_r_array_transp(ipoint,k)
fock_3_w_kl_wla_phi_k(ipoint,mm,a) += w_kl * w_la * mo_k
enddo
enddo
enddo
enddo
enddo
END_PROVIDER

View File

@ -1,287 +0,0 @@
! ---
BEGIN_PROVIDER [ double precision, two_e_vartc_integral_alpha, (ao_num, ao_num)]
&BEGIN_PROVIDER [ double precision, two_e_vartc_integral_beta , (ao_num, ao_num)]
implicit none
integer :: i, j, k, l
double precision :: density, density_a, density_b, I_coul, I_kjli
double precision :: t0, t1
double precision, allocatable :: tmp_a(:,:), tmp_b(:,:)
two_e_vartc_integral_alpha = 0.d0
two_e_vartc_integral_beta = 0.d0
!$OMP PARALLEL DEFAULT (NONE) &
!$OMP PRIVATE (i, j, k, l, density_a, density_b, density, tmp_a, tmp_b, I_coul, I_kjli) &
!$OMP SHARED (ao_num, TCSCF_density_matrix_ao_alpha, TCSCF_density_matrix_ao_beta, ao_two_e_tc_tot, &
!$OMP two_e_vartc_integral_alpha, two_e_vartc_integral_beta)
allocate(tmp_a(ao_num,ao_num), tmp_b(ao_num,ao_num))
tmp_a = 0.d0
tmp_b = 0.d0
!$OMP DO
do j = 1, ao_num
do l = 1, ao_num
density_a = TCSCF_density_matrix_ao_alpha(l,j)
density_b = TCSCF_density_matrix_ao_beta (l,j)
density = density_a + density_b
do i = 1, ao_num
do k = 1, ao_num
I_coul = density * ao_two_e_tc_tot(k,i,l,j)
I_kjli = ao_two_e_tc_tot(k,j,l,i)
tmp_a(k,i) += I_coul - density_a * I_kjli
tmp_b(k,i) += I_coul - density_b * I_kjli
enddo
enddo
enddo
enddo
!$OMP END DO NOWAIT
!$OMP CRITICAL
do i = 1, ao_num
do j = 1, ao_num
two_e_vartc_integral_alpha(j,i) += tmp_a(j,i)
two_e_vartc_integral_beta (j,i) += tmp_b(j,i)
enddo
enddo
!$OMP END CRITICAL
deallocate(tmp_a, tmp_b)
!$OMP END PARALLEL
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_ao_alpha, (ao_num, ao_num)]
implicit none
Fock_matrix_vartc_ao_alpha = ao_one_e_integrals_tc_tot + two_e_vartc_integral_alpha
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_ao_beta, (ao_num, ao_num)]
implicit none
Fock_matrix_vartc_ao_beta = ao_one_e_integrals_tc_tot + two_e_vartc_integral_beta
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_mo_alpha, (mo_num, mo_num) ]
implicit none
call ao_to_mo_bi_ortho( Fock_matrix_vartc_ao_alpha, size(Fock_matrix_vartc_ao_alpha, 1) &
, Fock_matrix_vartc_mo_alpha, size(Fock_matrix_vartc_mo_alpha, 1) )
if(three_body_h_tc) then
Fock_matrix_vartc_mo_alpha += fock_3e_uhf_mo_a
endif
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_mo_beta, (mo_num,mo_num) ]
implicit none
call ao_to_mo_bi_ortho( Fock_matrix_vartc_ao_beta, size(Fock_matrix_vartc_ao_beta, 1) &
, Fock_matrix_vartc_mo_beta, size(Fock_matrix_vartc_mo_beta, 1) )
if(three_body_h_tc) then
Fock_matrix_vartc_mo_beta += fock_3e_uhf_mo_b
endif
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, grad_vartc]
implicit none
integer :: i, k
double precision :: grad_left, grad_right
grad_left = 0.d0
grad_right = 0.d0
do i = 1, elec_beta_num ! doc --> SOMO
do k = elec_beta_num+1, elec_alpha_num
grad_left = max(grad_left , dabs(Fock_matrix_vartc_mo_tot(k,i)))
grad_right = max(grad_right, dabs(Fock_matrix_vartc_mo_tot(i,k)))
enddo
enddo
do i = 1, elec_beta_num ! doc --> virt
do k = elec_alpha_num+1, mo_num
grad_left = max(grad_left , dabs(Fock_matrix_vartc_mo_tot(k,i)))
grad_right = max(grad_right, dabs(Fock_matrix_vartc_mo_tot(i,k)))
enddo
enddo
do i = elec_beta_num+1, elec_alpha_num ! SOMO --> virt
do k = elec_alpha_num+1, mo_num
grad_left = max(grad_left , dabs(Fock_matrix_vartc_mo_tot(k,i)))
grad_right = max(grad_right, dabs(Fock_matrix_vartc_mo_tot(i,k)))
enddo
enddo
grad_vartc = grad_left + grad_right
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_ao_tot, (ao_num, ao_num) ]
implicit none
call mo_to_ao_bi_ortho( Fock_matrix_vartc_mo_tot, size(Fock_matrix_vartc_mo_tot, 1) &
, Fock_matrix_vartc_ao_tot, size(Fock_matrix_vartc_ao_tot, 1) )
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_mo_tot, (mo_num,mo_num) ]
&BEGIN_PROVIDER [ double precision, Fock_matrix_vartc_diag_mo_tot, (mo_num)]
implicit none
integer :: i, j, n
if(elec_alpha_num == elec_beta_num) then
Fock_matrix_vartc_mo_tot = Fock_matrix_vartc_mo_alpha
else
do j = 1, elec_beta_num
! F-K
do i = 1, elec_beta_num !CC
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))&
- (Fock_matrix_vartc_mo_beta(i,j) - Fock_matrix_vartc_mo_alpha(i,j))
enddo
! F+K/2
do i = elec_beta_num+1, elec_alpha_num !CA
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))&
+ 0.5d0*(Fock_matrix_vartc_mo_beta(i,j) - Fock_matrix_vartc_mo_alpha(i,j))
enddo
! F
do i = elec_alpha_num+1, mo_num !CV
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))
enddo
enddo
do j = elec_beta_num+1, elec_alpha_num
! F+K/2
do i = 1, elec_beta_num !AC
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))&
+ 0.5d0*(Fock_matrix_vartc_mo_beta(i,j) - Fock_matrix_vartc_mo_alpha(i,j))
enddo
! F
do i = elec_beta_num+1, elec_alpha_num !AA
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))
enddo
! F-K/2
do i = elec_alpha_num+1, mo_num !AV
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))&
- 0.5d0*(Fock_matrix_vartc_mo_beta(i,j) - Fock_matrix_vartc_mo_alpha(i,j))
enddo
enddo
do j = elec_alpha_num+1, mo_num
! F
do i = 1, elec_beta_num !VC
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))
enddo
! F-K/2
do i = elec_beta_num+1, elec_alpha_num !VA
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j))&
- 0.5d0*(Fock_matrix_vartc_mo_beta(i,j) - Fock_matrix_vartc_mo_alpha(i,j))
enddo
! F+K
do i = elec_alpha_num+1, mo_num !VV
Fock_matrix_vartc_mo_tot(i,j) = 0.5d0*(Fock_matrix_vartc_mo_alpha(i,j)+Fock_matrix_vartc_mo_beta(i,j)) &
+ (Fock_matrix_vartc_mo_beta(i,j) - Fock_matrix_vartc_mo_alpha(i,j))
enddo
enddo
if(three_body_h_tc)then
! C-O
do j = 1, elec_beta_num
do i = elec_beta_num+1, elec_alpha_num
Fock_matrix_vartc_mo_tot(i,j) += 0.5d0*(fock_a_tot_3e_bi_orth(i,j) + fock_b_tot_3e_bi_orth(i,j))
Fock_matrix_vartc_mo_tot(j,i) += 0.5d0*(fock_a_tot_3e_bi_orth(j,i) + fock_b_tot_3e_bi_orth(j,i))
enddo
enddo
! C-V
do j = 1, elec_beta_num
do i = elec_alpha_num+1, mo_num
Fock_matrix_vartc_mo_tot(i,j) += 0.5d0*(fock_a_tot_3e_bi_orth(i,j) + fock_b_tot_3e_bi_orth(i,j))
Fock_matrix_vartc_mo_tot(j,i) += 0.5d0*(fock_a_tot_3e_bi_orth(j,i) + fock_b_tot_3e_bi_orth(j,i))
enddo
enddo
! O-V
do j = elec_beta_num+1, elec_alpha_num
do i = elec_alpha_num+1, mo_num
Fock_matrix_vartc_mo_tot(i,j) += 0.5d0*(fock_a_tot_3e_bi_orth(i,j) + fock_b_tot_3e_bi_orth(i,j))
Fock_matrix_vartc_mo_tot(j,i) += 0.5d0*(fock_a_tot_3e_bi_orth(j,i) + fock_b_tot_3e_bi_orth(j,i))
enddo
enddo
endif
endif
do i = 1, mo_num
Fock_matrix_vartc_diag_mo_tot(i) = Fock_matrix_vartc_mo_tot(i,i)
enddo
if(frozen_orb_scf)then
integer :: iorb, jorb
do i = 1, n_core_orb
iorb = list_core(i)
do j = 1, n_act_orb
jorb = list_act(j)
Fock_matrix_vartc_mo_tot(iorb,jorb) = 0.d0
Fock_matrix_vartc_mo_tot(jorb,iorb) = 0.d0
enddo
enddo
endif
if(no_oa_or_av_opt)then
do i = 1, n_act_orb
iorb = list_act(i)
do j = 1, n_inact_orb
jorb = list_inact(j)
Fock_matrix_vartc_mo_tot(iorb,jorb) = 0.d0
Fock_matrix_vartc_mo_tot(jorb,iorb) = 0.d0
enddo
do j = 1, n_virt_orb
jorb = list_virt(j)
Fock_matrix_vartc_mo_tot(iorb,jorb) = 0.d0
Fock_matrix_vartc_mo_tot(jorb,iorb) = 0.d0
enddo
do j = 1, n_core_orb
jorb = list_core(j)
Fock_matrix_vartc_mo_tot(iorb,jorb) = 0.d0
Fock_matrix_vartc_mo_tot(jorb,iorb) = 0.d0
enddo
enddo
endif
!call check_sym(Fock_matrix_vartc_mo_tot, mo_num)
!do i = 1, mo_num
! write(*,'(100(F15.8, I4))') Fock_matrix_vartc_mo_tot(i,:)
!enddo
END_PROVIDER
! ---

View File

@ -1,391 +0,0 @@
! ---
BEGIN_PROVIDER [ double precision, tc_scf_dm_in_r, (n_points_final_grid) ]
implicit none
integer :: i, j
tc_scf_dm_in_r = 0.d0
do i = 1, n_points_final_grid
do j = 1, elec_beta_num
tc_scf_dm_in_r(i) += mos_r_in_r_array(j,i) * mos_l_in_r_array(j,i)
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, w_sum_in_r, (n_points_final_grid, 3)]
implicit none
integer :: ipoint, j, xi
w_sum_in_r = 0.d0
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
!w_sum_in_r(ipoint,xi) += x_W_ki_bi_ortho_erf_rk(ipoint,xi,j,j)
w_sum_in_r(ipoint,xi) += x_W_ki_bi_ortho_erf_rk_diag(ipoint,xi,j)
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, ww_sum_in_r, (n_points_final_grid, 3)]
implicit none
integer :: ipoint, j, xi
double precision :: tmp
ww_sum_in_r = 0.d0
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
tmp = x_W_ki_bi_ortho_erf_rk_diag(ipoint,xi,j)
ww_sum_in_r(ipoint,xi) += tmp * tmp
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, W1_r_in_r, (n_points_final_grid, 3, mo_num)]
implicit none
integer :: i, j, xi, ipoint
! TODO: call lapack
W1_r_in_r = 0.d0
do i = 1, mo_num
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
W1_r_in_r(ipoint,xi,i) += mos_r_in_r_array_transp(ipoint,j) * x_W_ki_bi_ortho_erf_rk(ipoint,xi,j,i)
enddo
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, W1_l_in_r, (n_points_final_grid, 3, mo_num)]
implicit none
integer :: i, j, xi, ipoint
! TODO: call lapack
W1_l_in_r = 0.d0
do i = 1, mo_num
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
W1_l_in_r(ipoint,xi,i) += mos_l_in_r_array_transp(ipoint,j) * x_W_ki_bi_ortho_erf_rk(ipoint,xi,i,j)
enddo
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, W1_in_r, (n_points_final_grid, 3)]
implicit none
integer :: j, xi, ipoint
! TODO: call lapack
W1_in_r = 0.d0
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
W1_in_r(ipoint,xi) += W1_l_in_r(ipoint,xi,j) * mos_r_in_r_array_transp(ipoint,j)
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, W1_diag_in_r, (n_points_final_grid, 3)]
implicit none
integer :: j, xi, ipoint
! TODO: call lapack
W1_diag_in_r = 0.d0
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
W1_diag_in_r(ipoint,xi) += mos_r_in_r_array_transp(ipoint,j) * mos_l_in_r_array_transp(ipoint,j) * x_W_ki_bi_ortho_erf_rk_diag(ipoint,xi,j)
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, v_sum_in_r, (n_points_final_grid, 3)]
implicit none
integer :: i, j, xi, ipoint
! TODO: call lapack
v_sum_in_r = 0.d0
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
v_sum_in_r(ipoint,xi) += x_W_ki_bi_ortho_erf_rk(ipoint,xi,i,j) * x_W_ki_bi_ortho_erf_rk(ipoint,xi,j,i)
enddo
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, W1_W1_r_in_r, (n_points_final_grid, 3, mo_num)]
implicit none
integer :: i, m, xi, ipoint
! TODO: call lapack
W1_W1_r_in_r = 0.d0
do i = 1, mo_num
do m = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
W1_W1_r_in_r(ipoint,xi,i) += x_W_ki_bi_ortho_erf_rk(ipoint,xi,m,i) * W1_r_in_r(ipoint,xi,m)
enddo
enddo
enddo
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, W1_W1_l_in_r, (n_points_final_grid, 3, mo_num)]
implicit none
integer :: i, j, xi, ipoint
! TODO: call lapack
W1_W1_l_in_r = 0.d0
do i = 1, mo_num
do j = 1, elec_beta_num
do xi = 1, 3
do ipoint = 1, n_points_final_grid
W1_W1_l_in_r(ipoint,xi,i) += x_W_ki_bi_ortho_erf_rk(ipoint,xi,i,j) * W1_l_in_r(ipoint,xi,j)
enddo
enddo
enddo
enddo
END_PROVIDER
! ---
subroutine direct_term_imj_bi_ortho(a, i, integral)
BEGIN_DOC
! computes sum_(j,m = 1, elec_beta_num) < a m j | i m j > with bi ortho mos
END_DOC
implicit none
integer, intent(in) :: i, a
double precision, intent(out) :: integral
integer :: ipoint, xi
double precision :: weight, tmp
integral = 0.d0
do xi = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
!integral += ( mos_l_in_r_array(a,ipoint) * mos_r_in_r_array(i,ipoint) * w_sum_in_r(ipoint,xi) * w_sum_in_r(ipoint,xi) &
! + 2.d0 * tc_scf_dm_in_r(ipoint) * w_sum_in_r(ipoint,xi) * x_W_ki_bi_ortho_erf_rk(ipoint,xi,a,i) ) * weight
tmp = w_sum_in_r(ipoint,xi)
integral += ( mos_l_in_r_array_transp(ipoint,a) * mos_r_in_r_array_transp(ipoint,i) * tmp * tmp &
+ 2.d0 * tc_scf_dm_in_r(ipoint) * tmp * x_W_ki_bi_ortho_erf_rk(ipoint,xi,a,i) &
) * weight
enddo
enddo
end
! ---
subroutine exch_term_jmi_bi_ortho(a, i, integral)
BEGIN_DOC
! computes sum_(j,m = 1, elec_beta_num) < a m j | j m i > with bi ortho mos
END_DOC
implicit none
integer, intent(in) :: i, a
double precision, intent(out) :: integral
integer :: ipoint, xi, j
double precision :: weight, tmp
integral = 0.d0
do xi = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
tmp = 0.d0
do j = 1, elec_beta_num
tmp = tmp + x_W_ki_bi_ortho_erf_rk(ipoint,xi,a,j) * x_W_ki_bi_ortho_erf_rk(ipoint,xi,j,i)
enddo
integral += ( mos_l_in_r_array_transp(ipoint,a) * W1_r_in_r(ipoint,xi,i) * w_sum_in_r(ipoint,xi) &
+ tc_scf_dm_in_r(ipoint) * tmp &
+ mos_r_in_r_array_transp(ipoint,i) * W1_l_in_r(ipoint,xi,a) * w_sum_in_r(ipoint,xi) &
) * weight
enddo
enddo
end
! ---
subroutine exch_term_ijm_bi_ortho(a, i, integral)
BEGIN_DOC
! computes sum_(j,m = 1, elec_beta_num) < a m j | i j m > with bi ortho mos
END_DOC
implicit none
integer, intent(in) :: i, a
double precision, intent(out) :: integral
integer :: ipoint, xi
double precision :: weight
integral = 0.d0
do xi = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
integral += ( mos_l_in_r_array_transp(ipoint,a) * mos_r_in_r_array_transp(ipoint,i) * v_sum_in_r(ipoint,xi) &
+ 2.d0 * x_W_ki_bi_ortho_erf_rk(ipoint,xi,a,i) * W1_in_r(ipoint,xi) &
) * weight
enddo
enddo
end
! ---
subroutine direct_term_ijj_bi_ortho(a, i, integral)
BEGIN_DOC
! computes sum_(j = 1, elec_beta_num) < a j j | i j j > with bi ortho mos
END_DOC
implicit none
integer, intent(in) :: i, a
double precision, intent(out) :: integral
integer :: ipoint, xi
double precision :: weight
integral = 0.d0
do xi = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
integral += ( mos_l_in_r_array_transp(ipoint,a) * mos_r_in_r_array_transp(ipoint,i) * ww_sum_in_r(ipoint,xi) &
+ 2.d0 * W1_diag_in_r(ipoint, xi) * x_W_ki_bi_ortho_erf_rk(ipoint,xi,a,i) &
) * weight
enddo
enddo
end
! ---
subroutine cyclic_term_jim_bi_ortho(a, i, integral)
BEGIN_DOC
! computes sum_(j,m = 1, elec_beta_num) < a m j | j i m > with bi ortho mos
END_DOC
implicit none
integer, intent(in) :: i, a
double precision, intent(out) :: integral
integer :: ipoint, xi
double precision :: weight
integral = 0.d0
do xi = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
integral += ( mos_l_in_r_array_transp(ipoint,a) * W1_W1_r_in_r(ipoint,xi,i) &
+ W1_W1_l_in_r(ipoint,xi,a) * mos_r_in_r_array_transp(ipoint,i) &
+ W1_l_in_r(ipoint,xi,a) * W1_r_in_r(ipoint,xi,i) &
) * weight
enddo
enddo
end
! ---
subroutine cyclic_term_mji_bi_ortho(a, i, integral)
BEGIN_DOC
! computes sum_(j,m = 1, elec_beta_num) < a m j | m j i > with bi ortho mos
END_DOC
implicit none
integer, intent(in) :: i, a
double precision, intent(out) :: integral
integer :: ipoint, xi
double precision :: weight
integral = 0.d0
do xi = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
integral += ( mos_l_in_r_array_transp(ipoint,a) * W1_W1_r_in_r(ipoint,xi,i) &
+ W1_l_in_r(ipoint,xi,a) * W1_r_in_r(ipoint,xi,i) &
+ W1_W1_l_in_r(ipoint,xi,a) * mos_r_in_r_array_transp(ipoint,i) &
) * weight
enddo
enddo
end
! ---

View File

@ -1,318 +0,0 @@
BEGIN_PROVIDER [integer , m_max_sm_7]
&BEGIN_PROVIDER [integer , n_max_sm_7]
&BEGIN_PROVIDER [integer , o_max_sm_7]
implicit none
BEGIN_DOC
! maximum value of the "m", "n" and "o" integer in the Jastrow function as in Eq. (4)
! of Schmidt,Moskowitz, JCP, 93, 4172 (1990) for the SM_7 version of Table IV
END_DOC
m_max_sm_7 = 4
n_max_sm_7 = 0
o_max_sm_7 = 4
END_PROVIDER
BEGIN_PROVIDER [integer , m_max_sm_9]
&BEGIN_PROVIDER [integer , n_max_sm_9]
&BEGIN_PROVIDER [integer , o_max_sm_9]
implicit none
BEGIN_DOC
! maximum value of the "m", "n" and "o" integer in the Jastrow function as in Eq. (4)
! of Schmidt,Moskowitz, JCP, 93, 4172 (1990) for the SM_9 version of Table IV
END_DOC
m_max_sm_9 = 4
n_max_sm_9 = 2
o_max_sm_9 = 4
END_PROVIDER
BEGIN_PROVIDER [integer , m_max_sm_17]
&BEGIN_PROVIDER [integer , n_max_sm_17]
&BEGIN_PROVIDER [integer , o_max_sm_17]
implicit none
BEGIN_DOC
! maximum value of the "m", "n" and "o" integer in the Jastrow function as in Eq. (4)
! of Schmidt,Moskowitz, JCP, 93, 4172 (1990) for the SM_17 version of Table IV
END_DOC
m_max_sm_17 = 6
n_max_sm_17 = 2
o_max_sm_17 = 6
END_PROVIDER
BEGIN_PROVIDER [ double precision, c_mn_o_sm_7, (0:m_max_sm_7,0:n_max_sm_7,0:o_max_sm_7,2:10)]
implicit none
BEGIN_DOC
!
!c_mn_o_7(0:4,0:4,2:10) = coefficient for the SM_7 correlation factor as given is Table IV of
! Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! the first index (0:4) is the "m" integer for the 1e part
! the second index(0:0) is the "n" integer for the 1e part WHICH IS ALWAYS SET TO 0 FOR SM_7
! the third index (0:4) is the "o" integer for the 2e part
! the fourth index (2:10) is the nuclear charge of the atom
END_DOC
c_mn_o_sm_7 = 0.d0
integer :: i
do i = 2, 10 ! loop over nuclear charge
c_mn_o_sm_7(0,0,1,i) = 0.5d0 ! all the linear terms are set to 1/2 to satisfy the anti-parallel spin condition
enddo
! He atom
! two electron terms
c_mn_o_sm_7(0,0,2,2) = 0.50516d0
c_mn_o_sm_7(0,0,3,2) = -0.19313d0
c_mn_o_sm_7(0,0,4,2) = 0.30276d0
! one-electron terms
c_mn_o_sm_7(2,0,0,2) = -0.16995d0
c_mn_o_sm_7(3,0,0,2) = -0.34505d0
c_mn_o_sm_7(4,0,0,2) = -0.54777d0
! Ne atom
! two electron terms
c_mn_o_sm_7(0,0,2,10) = -0.792d0
c_mn_o_sm_7(0,0,3,10) = 1.05232d0
c_mn_o_sm_7(0,0,4,10) = -0.65615d0
! one-electron terms
c_mn_o_sm_7(2,0,0,10) = -0.13312d0
c_mn_o_sm_7(3,0,0,10) = -0.00131d0
c_mn_o_sm_7(4,0,0,10) = 0.09083d0
END_PROVIDER
BEGIN_PROVIDER [ double precision, c_mn_o_sm_9, (0:m_max_sm_9,0:n_max_sm_9,0:o_max_sm_9,2:10)]
implicit none
BEGIN_DOC
!
!c_mn_o_9(0:4,0:4,2:10) = coefficient for the SM_9 correlation factor as given is Table IV of
! Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! the first index (0:4) is the "m" integer for the 1e part
! the second index(0:0) is the "n" integer for the 1e part WHICH IS ALWAYS SET TO 0 FOR SM_9
! the third index (0:4) is the "o" integer for the 2e part
! the fourth index (2:10) is the nuclear charge of the atom
END_DOC
c_mn_o_sm_9 = 0.d0
integer :: i
do i = 2, 10 ! loop over nuclear charge
c_mn_o_sm_9(0,0,1,i) = 0.5d0 ! all the linear terms are set to 1/2 to satisfy the anti-parallel spin condition
enddo
! He atom
! two electron terms
c_mn_o_sm_9(0,0,2,2) = 0.50516d0
c_mn_o_sm_9(0,0,3,2) = -0.19313d0
c_mn_o_sm_9(0,0,4,2) = 0.30276d0
! one-electron terms
c_mn_o_sm_9(2,0,0,2) = -0.16995d0
c_mn_o_sm_9(3,0,0,2) = -0.34505d0
c_mn_o_sm_9(4,0,0,2) = -0.54777d0
! Ne atom
! two electron terms
c_mn_o_sm_9(0,0,2,10) = -0.792d0
c_mn_o_sm_9(0,0,3,10) = 1.05232d0
c_mn_o_sm_9(0,0,4,10) = -0.65615d0
! one-electron terms
c_mn_o_sm_9(2,0,0,10) = -0.13312d0
c_mn_o_sm_9(3,0,0,10) = -0.00131d0
c_mn_o_sm_9(4,0,0,10) = 0.09083d0
END_PROVIDER
BEGIN_PROVIDER [ double precision, c_mn_o_sm_17, (0:m_max_sm_17,0:n_max_sm_17,0:o_max_sm_17,2:10)]
implicit none
BEGIN_DOC
!
!c_mn_o_17(0:4,0:4,2:10) = coefficient for the SM_17 correlation factor as given is Table IV of
! Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! the first index (0:4) is the "m" integer for the 1e part
! the second index(0:0) is the "n" integer for the 1e part WHICH IS ALWAYS SET TO 0 FOR SM_17
! the third index (0:4) is the "o" integer for the 2e part
! the fourth index (2:10) is the nuclear charge of the atom
END_DOC
c_mn_o_sm_17 = 0.d0
integer :: i
do i = 2, 10 ! loop over nuclear charge
c_mn_o_sm_17(0,0,1,i) = 0.5d0 ! all the linear terms are set to 1/2 to satisfy the anti-parallel spin condition
enddo
! He atom
! two electron terms
c_mn_o_sm_17(0,0,2,2) = 0.09239d0
c_mn_o_sm_17(0,0,3,2) = -0.38664d0
c_mn_o_sm_17(0,0,4,2) = 0.95764d0
! one-electron terms
c_mn_o_sm_17(2,0,0,2) = 0.23208d0
c_mn_o_sm_17(3,0,0,2) = -0.45032d0
c_mn_o_sm_17(4,0,0,2) = 0.82777d0
c_mn_o_sm_17(2,2,0,2) = -4.15388d0
! ee-n terms
c_mn_o_sm_17(2,0,2,2) = 0.80622d0
c_mn_o_sm_17(2,2,2,2) = 10.19704d0
c_mn_o_sm_17(4,0,2,2) = -4.96259d0
c_mn_o_sm_17(2,0,4,2) = -1.35647d0
c_mn_o_sm_17(4,2,2,2) = -5.90907d0
c_mn_o_sm_17(6,0,2,2) = 0.90343d0
c_mn_o_sm_17(4,0,4,2) = 5.50739d0
c_mn_o_sm_17(2,2,4,2) = -0.03154d0
c_mn_o_sm_17(2,0,6,2) = -1.1051860
! Ne atom
! two electron terms
c_mn_o_sm_17(0,0,2,10) = -0.80909d0
c_mn_o_sm_17(0,0,3,10) = -0.00219d0
c_mn_o_sm_17(0,0,4,10) = 0.59188d0
! one-electron terms
c_mn_o_sm_17(2,0,0,10) = -0.00567d0
c_mn_o_sm_17(3,0,0,10) = 0.14011d0
c_mn_o_sm_17(4,0,0,10) = -0.05671d0
c_mn_o_sm_17(2,2,0,10) = -3.33767d0
! ee-n terms
c_mn_o_sm_17(2,0,2,10) = 1.95067d0
c_mn_o_sm_17(2,2,2,10) = 6.83340d0
c_mn_o_sm_17(4,0,2,10) = -3.29231d0
c_mn_o_sm_17(2,0,4,10) = -2.44998d0
c_mn_o_sm_17(4,2,2,10) = -2.13029d0
c_mn_o_sm_17(6,0,2,10) = 2.25768d0
c_mn_o_sm_17(4,0,4,10) = 1.97951d0
c_mn_o_sm_17(2,2,4,10) = -2.0924160
c_mn_o_sm_17(2,0,6,10) = 0.35493d0
END_PROVIDER
BEGIN_PROVIDER [ double precision, b_I_sm_90,(2:10)]
&BEGIN_PROVIDER [ double precision, d_I_sm_90,(2:10)]
implicit none
BEGIN_DOC
! "b_I" and "d_I" parameters of Eqs. (4) and (5) of Schmidt,Moskowitz, JCP, 93, 4172 (1990)
END_DOC
b_I_sm_90 = 1.d0
d_I_sm_90 = 1.d0
END_PROVIDER
subroutine get_full_sm_90_jastrow(r1,r2,rI,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
implicit none
double precision, intent(in) :: r1(3),r2(3),rI(3)
integer, intent(in) :: sm_j, i_charge
double precision, intent(out):: j_1e,j_2e,j_een,j_tot
BEGIN_DOC
! Jastrow function as in Eq. (4) of Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! the i_charge variable is the integer specifying the charge of the atom for the Jastrow
! the sm_j integer variable represents the "quality" of the jastrow : sm_j = 7, 9, 17
END_DOC
double precision :: r_inucl,r_jnucl,r_ij,b_I, d_I
b_I = b_I_sm_90(i_charge)
d_I = d_I_sm_90(i_charge)
call get_rescaled_variables_j_sm_90(r1,r2,rI,b_I,d_I,r_inucl,r_jnucl,r_ij)
call jastrow_func_sm_90(r_inucl,r_jnucl,r_ij,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
end
subroutine get_rescaled_variables_j_sm_90(r1,r2,rI,b_I,d_I,r_inucl,r_jnucl,r_ij)
implicit none
BEGIN_DOC
! rescaled variables of Eq. (5) and (6) of Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! the "b_I" and "d_I" parameters are the same as in Eqs. (5) and (6)
END_DOC
double precision, intent(in) :: r1(3),r2(3),rI(3)
double precision, intent(in) :: b_I, d_I
double precision, intent(out):: r_inucl,r_jnucl,r_ij
double precision :: rin, rjn, rij
integer :: i
rin = 0.d0
rjn = 0.d0
rij = 0.d0
do i = 1,3
rin += (r1(i) - rI(i)) * (r1(i) - rI(i))
rjn += (r2(i) - rI(i)) * (r2(i) - rI(i))
rij += (r2(i) - r1(i)) * (r2(i) - r1(i))
enddo
rin = dsqrt(rin)
rjn = dsqrt(rjn)
rij = dsqrt(rij)
r_inucl = b_I * rin/(1.d0 + b_I * rin)
r_jnucl = b_I * rjn/(1.d0 + b_I * rjn)
r_ij = d_I * rij/(1.d0 + b_I * rij)
end
subroutine jastrow_func_sm_90(r_inucl,r_jnucl,r_ij,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
implicit none
BEGIN_DOC
! Jastrow function as in Eq. (4) of Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! Here the r_inucl, r_jnucl are the rescaled variables as defined in Eq. (5) with "b_I"
! r_ij is the rescaled variable as defined in Eq. (6) with "d_I"
! the i_charge variable is the integer specifying the charge of the atom for the Jastrow
! the sm_j integer variable represents the "quality" of the jastrow : sm_j = 7, 9, 17
!
! it returns the j_1e : sum of terms with "o" = "n" = 0, "m" /= 0,
! j_2e : sum of terms with "m" = "n" = 0, "o" /= 0,
! j_een : sum of terms with "m" /=0, "n" /= 0, "o" /= 0,
! j_tot : the total sum
END_DOC
double precision, intent(in) :: r_inucl,r_jnucl,r_ij
integer, intent(in) :: sm_j,i_charge
double precision, intent(out):: j_1e,j_2e,j_een,j_tot
j_1e = 0.D0
j_2e = 0.D0
j_een = 0.D0
double precision :: delta_mn,jastrow_sm_90_atomic
integer :: m,n,o
BEGIN_TEMPLATE
! pure 2e part
n = 0
m = 0
if(sm_j == $X )then
do o = 1, o_max_sm_$X
if(dabs(c_mn_o_sm_$X(m,n,o,i_charge)).lt.1.d-10)cycle
j_2e += c_mn_o_sm_$X(m,n,o,i_charge) * jastrow_sm_90_atomic(m,n,o,i_charge,r_inucl,r_jnucl,r_ij)
enddo
! else
! print*,'sm_j = ',sm_j
! print*,'not implemented, stop'
! stop
endif
! pure one-e part
o = 0
if(sm_j == $X)then
do n = 2, n_max_sm_$X
do m = 2, m_max_sm_$X
j_1e += c_mn_o_sm_$X(m,n,o,i_charge) * jastrow_sm_90_atomic(m,n,o,i_charge,r_inucl,r_jnucl,r_ij)
enddo
enddo
! else
! print*,'sm_j = ',sm_j
! print*,'not implemented, stop'
! stop
endif
! e-e-n part
if(sm_j == $X)then
do o = 1, o_max_sm_$X
do m = 2, m_max_sm_$X
do n = 2, n_max_sm_$X
j_een += c_mn_o_sm_$X(m,n,o,i_charge) * jastrow_sm_90_atomic(m,n,o,i_charge,r_inucl,r_jnucl,r_ij)
enddo
enddo
enddo
else
! print*,'sm_j = ',sm_j
! print*,'not implemented, stop'
! stop
endif
j_tot = j_1e + j_2e + j_een
SUBST [ X]
7 ;;
9 ;;
17 ;;
END_TEMPLATE
end
double precision function jastrow_sm_90_atomic(m,n,o,i_charge,r_inucl,r_jnucl,r_ij)
implicit none
BEGIN_DOC
! contribution to the function of Eq. (4) of Schmidt,Moskowitz, JCP, 93, 4172 (1990)
! for a given m,n,o and atom
END_DOC
double precision, intent(in) :: r_inucl,r_jnucl,r_ij
integer , intent(in) :: m,n,o,i_charge
double precision :: delta_mn
if(m==n)then
delta_mn = 0.5d0
else
delta_mn = 1.D0
endif
jastrow_sm_90_atomic = delta_mn * (r_inucl**m * r_jnucl**n + r_jnucl**m * r_inucl**n)*r_ij**o
end

View File

@ -1,69 +0,0 @@
program plot_j
implicit none
double precision :: r1(3),rI(3),r2(3)
double precision :: r12,dx,xmax, j_1e,j_2e,j_een,j_tot
double precision :: j_mu_F_x_j
integer :: i,nx,m,i_charge,sm_j
character*(128) :: output
integer :: i_unit_output_He_sm_7,i_unit_output_Ne_sm_7
integer :: i_unit_output_He_sm_17,i_unit_output_Ne_sm_17
integer :: getUnitAndOpen
output='J_SM_7_He'
i_unit_output_He_sm_7 = getUnitAndOpen(output,'w')
output='J_SM_7_Ne'
i_unit_output_Ne_sm_7 = getUnitAndOpen(output,'w')
output='J_SM_17_He'
i_unit_output_He_sm_17 = getUnitAndOpen(output,'w')
output='J_SM_17_Ne'
i_unit_output_Ne_sm_17 = getUnitAndOpen(output,'w')
rI = 0.d0
r1 = 0.d0
r2 = 0.d0
r1(1) = 1.5d0
xmax = 20.d0
r2(1) = -xmax*0.5d0
nx = 1000
dx = xmax/dble(nx)
do i = 1, nx
r12 = 0.d0
do m = 1, 3
r12 += (r1(m) - r2(m))*(r1(m) - r2(m))
enddo
r12 = dsqrt(r12)
double precision :: jmu,env_nucl,jmu_env,jmu_scaled, jmu_scaled_env
double precision :: b_I,d_I,r_inucl,r_jnucl,r_ij
b_I = 1.D0
d_I = 1.D0
call get_rescaled_variables_j_sm_90(r1,r2,rI,b_I,d_I,r_inucl,r_jnucl,r_ij)
jmu=j_mu_F_x_j(r12)
jmu_scaled=j_mu_F_x_j(r_ij)
jmu_env = jmu * env_nucl(r1) * env_nucl(r2)
! jmu_scaled_env= jmu_scaled * (1.d0 - env_coef(1) * dexp(-env_expo(1)*r_inucl**2)) * (1.d0 - env_coef(1) * dexp(-env_expo(1)*r_jnucl**2))
jmu_scaled_env= jmu_scaled * env_nucl(r1) * env_nucl(r2)
! He
i_charge = 2
! SM 7 Jastrow
sm_j = 7
call get_full_sm_90_jastrow(r1,r2,rI,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
write(i_unit_output_He_sm_7,'(100(F16.10,X))')r2(1),r12,j_mu_F_x_j(r12), j_1e,j_2e,j_een,j_tot,jmu_env,jmu_scaled,jmu_scaled_env
! SM 17 Jastrow
sm_j = 17
call get_full_sm_90_jastrow(r1,r2,rI,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
write(i_unit_output_He_sm_17,'(100(F16.10,X))')r2(1),r12,j_mu_F_x_j(r12), j_1e,j_2e,j_een,j_tot,jmu_env,jmu_scaled,jmu_scaled_env
! Ne
i_charge = 10
! SM 7 Jastrow
sm_j = 7
call get_full_sm_90_jastrow(r1,r2,rI,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
write(i_unit_output_Ne_sm_7,'(100(F16.10,X))')r2(1),r12,j_mu_F_x_j(r12), j_1e,j_2e,j_een,j_tot,jmu_env,jmu_scaled,jmu_scaled_env
! SM 17 Jastrow
sm_j = 17
call get_full_sm_90_jastrow(r1,r2,rI,sm_j,i_charge, j_1e,j_2e,j_een,j_tot)
write(i_unit_output_Ne_sm_17,'(100(F16.10,X))')r2(1),r12,j_mu_F_x_j(r12), j_1e,j_2e,j_een,j_tot,jmu_env,jmu_scaled,jmu_scaled_env
r2(1) += dx
enddo
end

View File

@ -1,59 +0,0 @@
program print_fit_param
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
implicit none
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
!call create_guess
!call orthonormalize_mos
call main()
end
! ---
subroutine main()
implicit none
integer :: i
mu_erf = 1.d0
touch mu_erf
print *, ' fit for (1 - erf(x))^2'
do i = 1, n_max_fit_slat
print*, expo_gauss_1_erf_x_2(i), coef_gauss_1_erf_x_2(i)
enddo
print *, ''
print *, ' fit for [x * (1 - erf(x)) - 1/sqrt(pi) * exp(-x**2)]'
do i = 1, n_max_fit_slat
print *, expo_gauss_j_mu_x(i), 2.d0 * coef_gauss_j_mu_x(i)
enddo
print *, ''
print *, ' fit for [x * (1 - erf(x)) - 1/sqrt(pi) * exp(-x**2)]^2'
do i = 1, n_max_fit_slat
print *, expo_gauss_j_mu_x_2(i), 4.d0 * coef_gauss_j_mu_x_2(i)
enddo
print *, ''
print *, ' fit for [x * (1 - erf(x)) - 1/sqrt(pi) * exp(-x**2)] x [1 - erf(mu * r12)]'
do i = 1, n_max_fit_slat
print *, expo_gauss_j_mu_1_erf(i), 4.d0 * coef_gauss_j_mu_1_erf(i)
enddo
return
end subroutine main
! ---

View File

@ -1,55 +0,0 @@
program print_tcscf_energy
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
implicit none
print *, 'Hello world'
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
call main()
end
! ---
subroutine main()
implicit none
double precision :: etc_tot, etc_1e, etc_2e, etc_3e
PROVIDE j2e_type mu_erf
PROVIDE j1e_type j1e_coef j1e_expo
PROVIDE env_type env_coef env_expo
print*, ' j2e_type = ', j2e_type
print*, ' j1e_type = ', j1e_type
print*, ' env_type = ', env_type
print*, ' mu_erf = ', mu_erf
etc_tot = TC_HF_energy
etc_1e = TC_HF_one_e_energy
etc_2e = TC_HF_two_e_energy
etc_3e = 0.d0
if(three_body_h_tc) then
!etc_3e = diag_three_elem_hf
etc_3e = tcscf_energy_3e_naive
endif
print *, " E_TC = ", etc_tot
print *, " E_1e = ", etc_1e
print *, " E_2e = ", etc_2e
print *, " E_3e = ", etc_3e
return
end subroutine main
! ---

View File

@ -61,7 +61,7 @@ subroutine rh_tcscf_diis()
etc_tot = TC_HF_energy
etc_1e = TC_HF_one_e_energy
etc_2e = TC_HF_two_e_energy
etc_3e = diag_three_elem_hf
etc_3e = TC_HF_three_e_energy
!tc_grad = grad_non_hermit
er_DIIS = maxval(abs(FQS_SQF_mo))
e_delta = dabs(etc_tot - e_save)
@ -189,7 +189,7 @@ subroutine rh_tcscf_diis()
etc_tot = TC_HF_energy
etc_1e = TC_HF_one_e_energy
etc_2e = TC_HF_two_e_energy
etc_3e = diag_three_elem_hf
etc_3e = TC_HF_three_e_energy
!tc_grad = grad_non_hermit
er_DIIS = maxval(abs(FQS_SQF_mo))
e_delta = dabs(etc_tot - e_save)
@ -234,7 +234,7 @@ subroutine rh_tcscf_diis()
call unlock_io
if(er_delta .lt. 0.d0) then
call ezfio_set_tc_scf_bitc_energy(etc_tot)
call ezfio_set_tc_scf_tcscf_energy(etc_tot)
call ezfio_set_bi_ortho_mos_mo_l_coef(mo_l_coef)
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)
write(json_unit, json_true_fmt) 'saved'
@ -263,7 +263,7 @@ subroutine rh_tcscf_diis()
deallocate(mo_r_coef_save, mo_l_coef_save, F_DIIS, E_DIIS)
call ezfio_set_tc_scf_bitc_energy(TC_HF_energy)
call ezfio_set_tc_scf_tcscf_energy(TC_HF_energy)
call ezfio_set_bi_ortho_mos_mo_l_coef(mo_l_coef)
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)

View File

@ -1,129 +0,0 @@
! ---
subroutine rh_tcscf_simple()
implicit none
integer :: i, j, it, dim_DIIS
double precision :: t0, t1
double precision :: e_save, e_delta, rho_delta
double precision :: etc_tot, etc_1e, etc_2e, etc_3e, tc_grad
double precision :: er_DIIS
double precision, allocatable :: rho_old(:,:), rho_new(:,:)
allocate(rho_old(ao_num,ao_num), rho_new(ao_num,ao_num))
it = 0
e_save = 0.d0
dim_DIIS = 0
! ---
if(.not. bi_ortho) then
print *, ' grad_hermit = ', grad_hermit
call save_good_hermit_tc_eigvectors
TOUCH mo_coef
call save_mos
endif
! ---
if(bi_ortho) then
PROVIDE level_shift_tcscf
PROVIDE mo_l_coef mo_r_coef
write(6, '(A4,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A4, 1X, A8)') &
'====', '================', '================', '================', '================', '================' &
, '================', '================', '================', '====', '========'
write(6, '(A4,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A4, 1X, A8)') &
' it ', ' SCF TC Energy ', ' E(1e) ', ' E(2e) ', ' E(3e) ', ' energy diff ' &
, ' gradient ', ' DIIS error ', ' level shift ', 'DIIS', ' WT (m)'
write(6, '(A4,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A4, 1X, A8)') &
'====', '================', '================', '================', '================', '================' &
, '================', '================', '================', '====', '========'
! first iteration (HF orbitals)
call wall_time(t0)
etc_tot = TC_HF_energy
etc_1e = TC_HF_one_e_energy
etc_2e = TC_HF_two_e_energy
etc_3e = 0.d0
if(three_body_h_tc) then
etc_3e = diag_three_elem_hf
endif
tc_grad = grad_non_hermit
er_DIIS = maxval(abs(FQS_SQF_mo))
e_delta = dabs(etc_tot - e_save)
e_save = etc_tot
call wall_time(t1)
write(6, '(I4,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, I4,1X, F8.2)') &
it, etc_tot, etc_1e, etc_2e, etc_3e, e_delta, tc_grad, er_DIIS, level_shift_tcscf, dim_DIIS, (t1-t0)/60.d0
do while(tc_grad .gt. dsqrt(thresh_tcscf))
call wall_time(t0)
it += 1
if(it > n_it_tcscf_max) then
print *, ' max of TCSCF iterations is reached ', n_it_TCSCF_max
stop
endif
mo_l_coef = fock_tc_leigvec_ao
mo_r_coef = fock_tc_reigvec_ao
call ezfio_set_bi_ortho_mos_mo_l_coef(mo_l_coef)
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)
TOUCH mo_l_coef mo_r_coef
etc_tot = TC_HF_energy
etc_1e = TC_HF_one_e_energy
etc_2e = TC_HF_two_e_energy
etc_3e = 0.d0
if(three_body_h_tc) then
etc_3e = diag_three_elem_hf
endif
tc_grad = grad_non_hermit
er_DIIS = maxval(abs(FQS_SQF_mo))
e_delta = dabs(etc_tot - e_save)
e_save = etc_tot
call ezfio_set_tc_scf_bitc_energy(etc_tot)
call wall_time(t1)
write(6, '(I4,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, I4,1X, F8.2)') &
it, etc_tot, etc_1e, etc_2e, etc_3e, e_delta, tc_grad, er_DIIS, level_shift_tcscf, dim_DIIS, (t1-t0)/60.d0
enddo
else
do while( (grad_hermit.gt.dsqrt(thresh_tcscf)) .and. (it.lt.n_it_tcscf_max) )
print*,'grad_hermit = ',grad_hermit
it += 1
print *, 'iteration = ', it
print *, '***'
print *, 'TC HF total energy = ', TC_HF_energy
print *, 'TC HF 1 e energy = ', TC_HF_one_e_energy
print *, 'TC HF 2 e energy = ', TC_HF_two_e_energy
print *, 'TC HF 3 body = ', diag_three_elem_hf
print *, '***'
print *, ''
call save_good_hermit_tc_eigvectors
TOUCH mo_coef
call save_mos
enddo
endif
print *, ' TCSCF Simple converged !'
!call print_energy_and_mos(good_angles)
deallocate(rho_old, rho_new)
end
! ---

View File

@ -1,89 +0,0 @@
! ---
subroutine rh_vartcscf_simple()
implicit none
integer :: i, j, it, dim_DIIS
double precision :: t0, t1
double precision :: e_save, e_delta, rho_delta
double precision :: etc_tot, etc_1e, etc_2e, etc_3e
double precision :: er_DIIS
it = 0
e_save = 0.d0
dim_DIIS = 0
! ---
PROVIDE level_shift_tcscf
PROVIDE mo_r_coef
write(6, '(A4,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A4, 1X, A8)') &
'====', '================', '================', '================', '================', '================' &
, '================', '================', '====', '========'
write(6, '(A4,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A4, 1X, A8)') &
' it ', ' SCF TC Energy ', ' E(1e) ', ' E(2e) ', ' E(3e) ', ' energy diff ' &
, ' DIIS error ', ' level shift ', 'DIIS', ' WT (m)'
write(6, '(A4,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A16,1X, A4, 1X, A8)') &
'====', '================', '================', '================', '================', '================' &
, '================', '================', '====', '========'
! first iteration (HF orbitals)
call wall_time(t0)
etc_tot = VARTC_HF_energy
etc_1e = VARTC_HF_one_e_energy
etc_2e = VARTC_HF_two_e_energy
etc_3e = 0.d0
if(three_body_h_tc) then
etc_3e = diag_three_elem_hf
endif
er_DIIS = maxval(abs(FQS_SQF_mo))
e_delta = dabs(etc_tot - e_save)
e_save = etc_tot
call wall_time(t1)
write(6, '(I4,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, I4,1X, F8.2)') &
it, etc_tot, etc_1e, etc_2e, etc_3e, e_delta, er_DIIS, level_shift_tcscf, dim_DIIS, (t1-t0)/60.d0
do while(er_DIIS .gt. dsqrt(thresh_tcscf))
call wall_time(t0)
it += 1
if(it > n_it_tcscf_max) then
print *, ' max of TCSCF iterations is reached ', n_it_TCSCF_max
stop
endif
mo_r_coef = fock_vartc_eigvec_ao
mo_l_coef = mo_r_coef
call ezfio_set_bi_ortho_mos_mo_l_coef(mo_l_coef)
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)
TOUCH mo_l_coef mo_r_coef
etc_tot = VARTC_HF_energy
etc_1e = VARTC_HF_one_e_energy
etc_2e = VARTC_HF_two_e_energy
etc_3e = 0.d0
if(three_body_h_tc) then
etc_3e = diag_three_elem_hf
endif
er_DIIS = maxval(abs(FQS_SQF_mo))
e_delta = dabs(etc_tot - e_save)
e_save = etc_tot
call ezfio_set_tc_scf_bitc_energy(etc_tot)
call wall_time(t1)
write(6, '(I4,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, F16.10,1X, I4,1X, F8.2)') &
it, etc_tot, etc_1e, etc_2e, etc_3e, e_delta, er_DIIS, level_shift_tcscf, dim_DIIS, (t1-t0)/60.d0
enddo
print *, ' VAR-TCSCF Simple converged !'
end
! ---

View File

@ -1,369 +0,0 @@
! ---
program rotate_tcscf_orbitals
BEGIN_DOC
! TODO : Rotate the bi-orthonormal orbitals in order to minimize left-right angles when degenerate
END_DOC
implicit none
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
bi_ortho = .True.
touch bi_ortho
call minimize_tc_orb_angles()
!call maximize_overlap()
end
! ---
subroutine maximize_overlap()
implicit none
integer :: i, m, n
double precision :: accu_d, accu_nd
double precision, allocatable :: C(:,:), R(:,:), L(:,:), W(:,:), e(:)
double precision, allocatable :: S(:,:)
n = ao_num
m = mo_num
allocate(L(n,m), R(n,m), C(n,m), W(n,n), e(m))
L = mo_l_coef
R = mo_r_coef
C = mo_coef
W = ao_overlap
print*, ' fock matrix diag elements'
do i = 1, m
e(i) = Fock_matrix_tc_mo_tot(i,i)
print*, e(i)
enddo
! ---
print *, ' overlap before :'
print *, ' '
allocate(S(m,m))
call LTxSxR(n, m, L, W, R, S)
!print*, " L.T x R"
!do i = 1, m
! write(*, '(100(F16.10,X))') S(i,i)
!enddo
call LTxSxR(n, m, L, W, C, S)
print*, " L.T x C"
do i = 1, m
write(*, '(100(F16.10,X))') S(i,:)
enddo
call LTxSxR(n, m, C, W, R, S)
print*, " C.T x R"
do i = 1, m
write(*, '(100(F16.10,X))') S(i,:)
enddo
deallocate(S)
! ---
call rotate_degen_eigvec_to_maximize_overlap(n, m, e, C, W, L, R)
! ---
print *, ' overlap after :'
print *, ' '
allocate(S(m,m))
call LTxSxR(n, m, L, W, R, S)
!print*, " L.T x R"
!do i = 1, m
! write(*, '(100(F16.10,X))') S(i,i)
!enddo
call LTxSxR(n, m, L, W, C, S)
print*, " L.T x C"
do i = 1, m
write(*, '(100(F16.10,X))') S(i,:)
enddo
call LTxSxR(n, m, C, W, R, S)
print*, " C.T x R"
do i = 1, m
write(*, '(100(F16.10,X))') S(i,:)
enddo
deallocate(S)
! ---
mo_l_coef = L
mo_r_coef = R
call ezfio_set_bi_ortho_mos_mo_l_coef(mo_l_coef)
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)
! ---
deallocate(L, R, C, W, e)
end subroutine maximize_overlap
! ---
subroutine rotate_degen_eigvec_to_maximize_overlap(n, m, e0, C0, W0, L0, R0)
implicit none
integer, intent(in) :: n, m
double precision, intent(in) :: e0(m), W0(n,n), C0(n,m)
double precision, intent(inout) :: L0(n,m), R0(n,m)
integer :: i, j, k, kk, mm, id1, tot_deg
double precision :: ei, ej, de, de_thr
integer, allocatable :: deg_num(:)
double precision, allocatable :: L(:,:), R(:,:), C(:,:), Lnew(:,:), Rnew(:,:), tmp(:,:)
!double precision, allocatable :: S(:,:), Snew(:,:), T(:,:), Ttmp(:,:), Stmp(:,:)
double precision, allocatable :: S(:,:), Snew(:,:), T(:,:), Ttmp(:,:), Stmp(:,:)
!real*8 :: S(m,m), Snew(m,m), T(m,m)
id1 = 700
allocate(S(id1,id1), Snew(id1,id1), T(id1,id1))
! ---
allocate( deg_num(m) )
do i = 1, m
deg_num(i) = 1
enddo
de_thr = thr_degen_tc
do i = 1, m-1
ei = e0(i)
! already considered in degen vectors
if(deg_num(i).eq.0) cycle
do j = i+1, m
ej = e0(j)
de = dabs(ei - ej)
if(de .lt. de_thr) then
deg_num(i) = deg_num(i) + 1
deg_num(j) = 0
endif
enddo
enddo
tot_deg = 0
do i = 1, m
if(deg_num(i).gt.1) then
print *, ' degen on', i, deg_num(i)
tot_deg = tot_deg + 1
endif
enddo
if(tot_deg .eq. 0) then
print *, ' no degen'
return
endif
! ---
do i = 1, m
mm = deg_num(i)
if(mm .gt. 1) then
allocate(L(n,mm), R(n,mm), C(n,mm))
do j = 1, mm
L(1:n,j) = L0(1:n,i+j-1)
R(1:n,j) = R0(1:n,i+j-1)
C(1:n,j) = C0(1:n,i+j-1)
enddo
! ---
! C.T x W0 x R
allocate(tmp(mm,n), Stmp(mm,mm))
call dgemm( 'T', 'N', mm, n, n, 1.d0 &
, C, size(C, 1), W0, size(W0, 1) &
, 0.d0, tmp, size(tmp, 1) )
call dgemm( 'N', 'N', mm, mm, n, 1.d0 &
, tmp, size(tmp, 1), R, size(R, 1) &
, 0.d0, Stmp, size(Stmp, 1) )
deallocate(C, tmp)
S = 0.d0
do k = 1, mm
do kk = 1, mm
S(kk,k) = Stmp(kk,k)
enddo
enddo
deallocate(Stmp)
!print*, " overlap bef"
!do k = 1, mm
! write(*, '(100(F16.10,X))') (S(k,kk), kk=1, mm)
!enddo
T = 0.d0
Snew = 0.d0
call maxovl(mm, mm, S, T, Snew)
!print*, " overlap aft"
!do k = 1, mm
! write(*, '(100(F16.10,X))') (Snew(k,kk), kk=1, mm)
!enddo
allocate(Ttmp(mm,mm))
Ttmp(1:mm,1:mm) = T(1:mm,1:mm)
allocate(Lnew(n,mm), Rnew(n,mm))
call dgemm( 'N', 'N', n, mm, mm, 1.d0 &
, R, size(R, 1), Ttmp(1,1), size(Ttmp, 1) &
, 0.d0, Rnew, size(Rnew, 1) )
call dgemm( 'N', 'N', n, mm, mm, 1.d0 &
, L, size(L, 1), Ttmp(1,1), size(Ttmp, 1) &
, 0.d0, Lnew, size(Lnew, 1) )
deallocate(L, R)
deallocate(Ttmp)
! ---
do j = 1, mm
L0(1:n,i+j-1) = Lnew(1:n,j)
R0(1:n,i+j-1) = Rnew(1:n,j)
enddo
deallocate(Lnew, Rnew)
endif
enddo
deallocate(S, Snew, T)
end subroutine rotate_degen_eigvec_to_maximize_overlap
! ---
subroutine fix_right_to_one()
implicit none
integer :: i, j, m, n, mm, tot_deg
double precision :: accu_d, accu_nd
double precision :: de_thr, ei, ej, de
integer, allocatable :: deg_num(:)
double precision, allocatable :: R0(:,:), L0(:,:), W(:,:), e0(:)
double precision, allocatable :: R(:,:), L(:,:), S(:,:), Stmp(:,:), tmp(:,:)
n = ao_num
m = mo_num
allocate(L0(n,m), R0(n,m), W(n,n), e0(m))
L0 = mo_l_coef
R0 = mo_r_coef
W = ao_overlap
print*, ' fock matrix diag elements'
do i = 1, m
e0(i) = Fock_matrix_tc_mo_tot(i,i)
print*, e0(i)
enddo
! ---
allocate( deg_num(m) )
do i = 1, m
deg_num(i) = 1
enddo
de_thr = 1d-6
do i = 1, m-1
ei = e0(i)
! already considered in degen vectors
if(deg_num(i).eq.0) cycle
do j = i+1, m
ej = e0(j)
de = dabs(ei - ej)
if(de .lt. de_thr) then
deg_num(i) = deg_num(i) + 1
deg_num(j) = 0
endif
enddo
enddo
deallocate(e0)
tot_deg = 0
do i = 1, m
if(deg_num(i).gt.1) then
print *, ' degen on', i, deg_num(i)
tot_deg = tot_deg + 1
endif
enddo
if(tot_deg .eq. 0) then
print *, ' no degen'
return
endif
! ---
do i = 1, m
mm = deg_num(i)
if(mm .gt. 1) then
allocate(L(n,mm), R(n,mm))
do j = 1, mm
L(1:n,j) = L0(1:n,i+j-1)
R(1:n,j) = R0(1:n,i+j-1)
enddo
! ---
call impose_weighted_orthog_svd(n, mm, W, R)
call impose_weighted_biorthog_qr(n, mm, thresh_biorthog_diag, thresh_biorthog_nondiag, R, W, L)
! ---
do j = 1, mm
L0(1:n,i+j-1) = L(1:n,j)
R0(1:n,i+j-1) = R(1:n,j)
enddo
deallocate(L, R)
endif
enddo
call check_weighted_biorthog_binormalize(n, m, L0, W, R0, thresh_biorthog_diag, thresh_biorthog_nondiag, .true.)
deallocate(W, deg_num)
mo_l_coef = L0
mo_r_coef = R0
deallocate(L0, R0)
call ezfio_set_bi_ortho_mos_mo_l_coef(mo_l_coef)
call ezfio_set_bi_ortho_mos_mo_r_coef(mo_r_coef)
print *, ' orbitals are rotated '
return
end subroutine fix_right_to_one
! ---

View File

@ -1,91 +0,0 @@
! ---
program tc_petermann_factor
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
implicit none
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
call main()
end
! ---
subroutine main()
implicit none
integer :: i, j
double precision :: Pf_diag_av
double precision, allocatable :: Sl(:,:), Sr(:,:), Pf(:,:)
allocate(Sl(mo_num,mo_num), Sr(mo_num,mo_num), Pf(mo_num,mo_num))
call LTxSxR(ao_num, mo_num, mo_l_coef, ao_overlap, mo_r_coef, Sl)
!call dgemm( "T", "N", mo_num, mo_num, ao_num, 1.d0 &
! , mo_l_coef, size(mo_l_coef, 1), mo_l_coef, size(mo_l_coef, 1) &
! , 0.d0, Sl, size(Sl, 1) )
print *, ''
print *, ' left-right orthog matrix:'
do i = 1, mo_num
write(*,'(100(F8.4,X))') Sl(:,i)
enddo
call LTxSxR(ao_num, mo_num, mo_l_coef, ao_overlap, mo_l_coef, Sl)
!call dgemm( "T", "N", mo_num, mo_num, ao_num, 1.d0 &
! , mo_l_coef, size(mo_l_coef, 1), mo_l_coef, size(mo_l_coef, 1) &
! , 0.d0, Sl, size(Sl, 1) )
print *, ''
print *, ' left-orthog matrix:'
do i = 1, mo_num
write(*,'(100(F8.4,X))') Sl(:,i)
enddo
call LTxSxR(ao_num, mo_num, mo_r_coef, ao_overlap, mo_r_coef, Sr)
! call dgemm( "T", "N", mo_num, mo_num, ao_num, 1.d0 &
! , mo_r_coef, size(mo_r_coef, 1), mo_r_coef, size(mo_r_coef, 1) &
! , 0.d0, Sr, size(Sr, 1) )
print *, ''
print *, ' right-orthog matrix:'
do i = 1, mo_num
write(*,'(100(F8.4,X))') Sr(:,i)
enddo
print *, ''
print *, ' Petermann matrix:'
do i = 1, mo_num
do j = 1, mo_num
Pf(j,i) = Sl(j,i) * Sr(j,i)
enddo
write(*,'(100(F8.4,X))') Pf(:,i)
enddo
Pf_diag_av = 0.d0
do i = 1, mo_num
Pf_diag_av = Pf_diag_av + Pf(i,i)
enddo
Pf_diag_av = Pf_diag_av / dble(mo_num)
print *, ''
print *, ' mean of the diagonal Petermann factor = ', Pf_diag_av
deallocate(Sl, Sr, Pf)
return
end subroutine
! ---

View File

@ -7,19 +7,6 @@ program tc_scf
END_DOC
implicit none
integer :: i
logical :: good_angles
PROVIDE j1e_type
PROVIDE j2e_type
PROVIDE tcscf_algorithm
PROVIDE var_tc
print *, ' TC-SCF with:'
print *, ' j1e_type = ', j1e_type
print *, ' j2e_type = ', j2e_type
write(json_unit,json_array_open_fmt) 'tc-scf'
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
@ -30,7 +17,6 @@ program tc_scf
call write_int(6, my_n_pt_r_grid, 'radial external grid over')
call write_int(6, my_n_pt_a_grid, 'angular external grid over')
if(tc_integ_type .eq. "numeric") then
my_extra_grid_becke = .True.
PROVIDE tc_grid2_a tc_grid2_r
@ -42,48 +28,38 @@ program tc_scf
call write_int(6, my_n_pt_a_extra_grid, 'angular internal grid over')
endif
!call create_guess()
!call orthonormalize_mos()
call main()
end
if(var_tc) then
! ---
print *, ' VAR-TC'
subroutine main()
if(tcscf_algorithm == 'DIIS') then
print*, ' NOT implemented yet'
elseif(tcscf_algorithm == 'Simple') then
call rh_vartcscf_simple()
else
print *, ' not implemented yet', tcscf_algorithm
stop
endif
implicit none
else
integer :: i
logical :: good_angles
if(tcscf_algorithm == 'DIIS') then
call rh_tcscf_diis()
elseif(tcscf_algorithm == 'Simple') then
call rh_tcscf_simple()
else
print *, ' not implemented yet', tcscf_algorithm
stop
endif
print *, ' TC-SCF with:'
print *, ' j2e_type = ', j2e_type
print *, ' j1e_type = ', j1e_type
print *, ' env_type = ', env_type
PROVIDE Fock_matrix_tc_diag_mo_tot
print*, ' Eigenvalues:'
do i = 1, mo_num
print*, i, Fock_matrix_tc_diag_mo_tot(i)
enddo
write(json_unit,json_array_open_fmt) 'tc-scf'
! TODO
! rotate angles in separate code only if necessary
if(minimize_lr_angles)then
call minimize_tc_orb_angles()
endif
call print_energy_and_mos(good_angles)
call rh_tcscf_diis()
PROVIDE Fock_matrix_tc_diag_mo_tot
print*, ' Eigenvalues:'
do i = 1, mo_num
print*, i, Fock_matrix_tc_diag_mo_tot(i)
enddo
if(minimize_lr_angles) then
call minimize_tc_orb_angles()
endif
call print_energy_and_mos(good_angles)
write(json_unit,json_array_close_fmtx)
call json_close
@ -119,7 +95,7 @@ subroutine create_guess()
SOFT_TOUCH mo_label
endif
end subroutine create_guess
end
! ---

View File

@ -10,16 +10,8 @@ BEGIN_PROVIDER [double precision, TCSCF_density_matrix_ao_beta, (ao_num, ao_num)
implicit none
if(bi_ortho) then
PROVIDE mo_l_coef mo_r_coef
TCSCF_density_matrix_ao_beta = TCSCF_bi_ort_dm_ao_beta
else
TCSCF_density_matrix_ao_beta = SCF_density_matrix_ao_beta
endif
PROVIDE mo_l_coef mo_r_coef
TCSCF_density_matrix_ao_beta = TCSCF_bi_ort_dm_ao_beta
END_PROVIDER
@ -35,16 +27,8 @@ BEGIN_PROVIDER [double precision, TCSCF_density_matrix_ao_alpha, (ao_num, ao_num
implicit none
if(bi_ortho) then
PROVIDE mo_l_coef mo_r_coef
TCSCF_density_matrix_ao_alpha = TCSCF_bi_ort_dm_ao_alpha
else
TCSCF_density_matrix_ao_alpha = SCF_density_matrix_ao_alpha
endif
PROVIDE mo_l_coef mo_r_coef
TCSCF_density_matrix_ao_alpha = TCSCF_bi_ort_dm_ao_alpha
END_PROVIDER

View File

@ -1,7 +1,8 @@
BEGIN_PROVIDER [ double precision, TC_HF_energy ]
&BEGIN_PROVIDER [ double precision, TC_HF_one_e_energy]
&BEGIN_PROVIDER [ double precision, TC_HF_two_e_energy]
BEGIN_PROVIDER [double precision, TC_HF_energy ]
&BEGIN_PROVIDER [double precision, TC_HF_one_e_energy ]
&BEGIN_PROVIDER [double precision, TC_HF_two_e_energy ]
&BEGIN_PROVIDER [double precision, TC_HF_three_e_energy]
BEGIN_DOC
! TC Hartree-Fock energy containing the nuclear repulsion, and its one- and two-body components.
@ -11,11 +12,8 @@
integer :: i, j
double precision :: t0, t1
!print*, ' Providing TC energy ...'
!call wall_time(t0)
PROVIDE mo_l_coef mo_r_coef
PROVIDE two_e_tc_non_hermit_integral_alpha two_e_tc_non_hermit_integral_beta
PROVIDE two_e_tc_integral_alpha two_e_tc_integral_beta
TC_HF_energy = nuclear_repulsion
TC_HF_one_e_energy = 0.d0
@ -23,47 +21,20 @@
do j = 1, ao_num
do i = 1, ao_num
TC_HF_two_e_energy += 0.5d0 * ( two_e_tc_non_hermit_integral_alpha(i,j) * TCSCF_density_matrix_ao_alpha(i,j) &
+ two_e_tc_non_hermit_integral_beta (i,j) * TCSCF_density_matrix_ao_beta (i,j) )
TC_HF_two_e_energy += 0.5d0 * ( two_e_tc_integral_alpha(i,j) * TCSCF_density_matrix_ao_alpha(i,j) &
+ two_e_tc_integral_beta (i,j) * TCSCF_density_matrix_ao_beta (i,j) )
TC_HF_one_e_energy += ao_one_e_integrals_tc_tot(i,j) &
* (TCSCF_density_matrix_ao_alpha(i,j) + TCSCF_density_matrix_ao_beta (i,j) )
enddo
enddo
TC_HF_energy += TC_HF_one_e_energy + TC_HF_two_e_energy
TC_HF_energy += diag_three_elem_hf
if((three_body_h_tc .eq. .False.) .and. (.not. noL_standard)) then
TC_HF_three_e_energy = 0.d0
else
TC_HF_three_e_energy = noL_0e
endif
!call wall_time(t1)
!print*, ' Wall time for TC energy=', t1-t0
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, VARTC_HF_energy]
&BEGIN_PROVIDER [ double precision, VARTC_HF_one_e_energy]
&BEGIN_PROVIDER [ double precision, VARTC_HF_two_e_energy]
implicit none
integer :: i, j
PROVIDE mo_r_coef
VARTC_HF_energy = nuclear_repulsion
VARTC_HF_one_e_energy = 0.d0
VARTC_HF_two_e_energy = 0.d0
do j = 1, ao_num
do i = 1, ao_num
VARTC_HF_two_e_energy += 0.5d0 * ( two_e_vartc_integral_alpha(i,j) * TCSCF_density_matrix_ao_alpha(i,j) &
+ two_e_vartc_integral_beta (i,j) * TCSCF_density_matrix_ao_beta (i,j) )
VARTC_HF_one_e_energy += ao_one_e_integrals_tc_tot(i,j) &
* (TCSCF_density_matrix_ao_alpha(i,j) + TCSCF_density_matrix_ao_beta (i,j) )
enddo
enddo
VARTC_HF_energy += VARTC_HF_one_e_energy + VARTC_HF_two_e_energy
VARTC_HF_energy += diag_three_elem_hf
TC_HF_energy += TC_HF_one_e_energy + TC_HF_two_e_energy + TC_HF_three_e_energy
END_PROVIDER

View File

@ -1,80 +0,0 @@
! ---
BEGIN_PROVIDER [double precision, tcscf_energy_3e_naive]
implicit none
integer :: i, j, k
integer :: neu, ned, D(elec_num)
integer :: ii, jj, kk
integer :: si, sj, sk
double precision :: I_ijk, I_jki, I_kij, I_jik, I_ikj, I_kji
double precision :: I_tot
PROVIDE mo_l_coef mo_r_coef
neu = elec_alpha_num
ned = elec_beta_num
if (neu > 0) D(1:neu) = [(2*i-1, i = 1, neu)]
if (ned > 0) D(neu+1:neu+ned) = [(2*i, i = 1, ned)]
!print*, "D = "
!do i = 1, elec_num
! ii = (D(i) - 1) / 2 + 1
! si = mod(D(i), 2)
! print*, i, D(i), ii, si
!enddo
tcscf_energy_3e_naive = 0.d0
do i = 1, elec_num - 2
ii = (D(i) - 1) / 2 + 1
si = mod(D(i), 2)
do j = i + 1, elec_num - 1
jj = (D(j) - 1) / 2 + 1
sj = mod(D(j), 2)
do k = j + 1, elec_num
kk = (D(k) - 1) / 2 + 1
sk = mod(D(k), 2)
call give_integrals_3_body_bi_ort(ii, jj, kk, ii, jj, kk, I_ijk)
I_tot = I_ijk
if(sj==si .and. sk==sj) then
call give_integrals_3_body_bi_ort(ii, jj, kk, jj, kk, ii, I_jki)
I_tot += I_jki
endif
if(sk==si .and. si==sj) then
call give_integrals_3_body_bi_ort(ii, jj, kk, kk, ii, jj, I_kij)
I_tot += I_kij
endif
if(sj==si) then
call give_integrals_3_body_bi_ort(ii, jj, kk, jj, ii, kk, I_jik)
I_tot -= I_jik
endif
if(sk==sj) then
call give_integrals_3_body_bi_ort(ii, jj, kk, ii, kk, jj, I_ikj)
I_tot -= I_ikj
endif
if(sk==si) then
call give_integrals_3_body_bi_ort(ii, jj, kk, kk, jj, ii, I_kji)
I_tot -= I_kji
endif
tcscf_energy_3e_naive += I_tot
enddo
enddo
enddo
tcscf_energy_3e_naive = -tcscf_energy_3e_naive
END_PROVIDER
! ---

View File

@ -1,970 +0,0 @@
program test_ints
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
implicit none
print *, ' starting test_ints ...'
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r
my_n_pt_r_grid = tc_grid1_r
my_n_pt_a_grid = tc_grid1_a
touch my_grid_becke my_n_pt_r_grid my_n_pt_a_grid
my_extra_grid_becke = .True.
my_n_pt_r_extra_grid = 30
my_n_pt_a_extra_grid = 50 ! small extra_grid for quick debug
touch my_extra_grid_becke my_n_pt_r_extra_grid my_n_pt_a_extra_grid
!! OK
! call routine_int2_u_grad1u_env2
! OK
! call routine_v_ij_erf_rk_cst_mu_env
! OK
! call routine_x_v_ij_erf_rk_cst_mu_env
! OK
! call routine_int2_u2_env2
! OK
! call routine_int2_u_grad1u_x_env2
! OK
! call routine_int2_grad1u2_grad2u2_env2
! call routine_int2_u_grad1u_env2
! call test_int2_grad1_u12_ao_test
! call routine_v_ij_u_cst_mu_env_test
! call test_grid_points_ao
!call test_int_gauss
!call test_fock_3e_uhf_ao()
!call test_fock_3e_uhf_mo()
!call test_two_e_tc_non_hermit_integral()
!!PROVIDE TC_HF_energy VARTC_HF_energy
!!print *, ' TC_HF_energy = ', TC_HF_energy
!!print *, ' VARTC_HF_energy = ', VARTC_HF_energy
call test_fock_3e_uhf_mo_cs()
call test_fock_3e_uhf_mo_a()
call test_fock_3e_uhf_mo_b()
end
! ---
subroutine routine_test_env
implicit none
integer :: i,icount,j
icount = 0
do i = 1, List_env1s_square_size
if(dabs(List_env1s_square_coef(i)).gt.1.d-10)then
print*,''
print*,List_env1s_square_expo(i),List_env1s_square_coef(i)
print*,List_env1s_square_cent(1:3,i)
print*,''
icount += 1
endif
enddo
print*,'List_env1s_square_coef,icount = ',List_env1s_square_size,icount
do i = 1, ao_num
do j = 1, ao_num
do icount = 1, List_comb_thr_b3_size(j,i)
print*,'',j,i
print*,List_comb_thr_b3_expo(icount,j,i),List_comb_thr_b3_coef(icount,j,i)
print*,List_comb_thr_b3_cent(1:3,icount,j,i)
print*,''
enddo
! enddo
enddo
enddo
print*,'max_List_comb_thr_b3_size = ',max_List_comb_thr_b3_size,List_env1s_square_size
end
subroutine routine_int2_u_grad1u_env2
implicit none
integer :: i,j,ipoint,k,l
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += int2_u_grad1u_env2_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += int2_u_grad1u_env2(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_int2_u_grad1u_env2'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_v_ij_erf_rk_cst_mu_env
implicit none
integer :: i,j,ipoint,k,l
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += v_ij_erf_rk_cst_mu_env_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += v_ij_erf_rk_cst_mu_env(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_v_ij_erf_rk_cst_mu_env'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_x_v_ij_erf_rk_cst_mu_env
implicit none
integer :: i,j,ipoint,k,l,m
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
do m = 1, 3
array(j,i,l,k) += x_v_ij_erf_rk_cst_mu_env_test(j,i,ipoint,m) * aos_grad_in_r_array_transp(m,k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += x_v_ij_erf_rk_cst_mu_env (j,i,ipoint,m) * aos_grad_in_r_array_transp(m,k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_x_v_ij_erf_rk_cst_mu_env'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_v_ij_u_cst_mu_env_test
implicit none
integer :: i,j,ipoint,k,l
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += v_ij_u_cst_mu_env_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += v_ij_u_cst_mu_env_fit (j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_v_ij_u_cst_mu_env_test'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_int2_grad1u2_grad2u2_env2
implicit none
integer :: i,j,ipoint,k,l
integer :: ii , jj
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
double precision, allocatable :: ints(:,:,:)
allocate(ints(ao_num, ao_num, n_points_final_grid))
! do ipoint = 1, n_points_final_grid
! do i = 1, ao_num
! do j = 1, ao_num
! read(33,*)ints(j,i,ipoint)
! enddo
! enddo
! enddo
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += int2_grad1u2_grad2u2_env2_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
! !array(j,i,l,k) += int2_grad1u2_grad2u2_env2_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
! array_ref(j,i,l,k) += int2_grad1u2_grad2u2_env2_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
! !array(j,i,l,k) += ints(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
! array_ref(j,i,l,k) += int2_grad1u2_grad2u2_env2(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += ints(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
! if(dabs(int2_grad1u2_grad2u2_env2_test(j,i,ipoint)).gt.1.d-6)then
! if(dabs(int2_grad1u2_grad2u2_env2_test(j,i,ipoint) - int2_grad1u2_grad2u2_env2_test(j,i,ipoint)).gt.1.d-6)then
! print*,j,i,ipoint
! print*,int2_grad1u2_grad2u2_env2_test(j,i,ipoint) , int2_grad1u2_grad2u2_env2_test(j,i,ipoint), dabs(int2_grad1u2_grad2u2_env2_test(j,i,ipoint) - int2_grad1u2_grad2u2_env2_test(j,i,ipoint))
! print*,int2_grad1u2_grad2u2_env2_test(i,j,ipoint) , int2_grad1u2_grad2u2_env2_test(i,j,ipoint), dabs(int2_grad1u2_grad2u2_env2_test(i,j,ipoint) - int2_grad1u2_grad2u2_env2_test(i,j,ipoint))
! stop
! endif
! endif
enddo
enddo
enddo
enddo
enddo
double precision :: e_ref, e_new
accu_relat = 0.d0
accu_abs = 0.d0
e_ref = 0.d0
e_new = 0.d0
do ii = 1, elec_alpha_num
do jj = ii, elec_alpha_num
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
e_ref += mo_coef(j,ii) * mo_coef(i,ii) * array_ref(j,i,l,k) * mo_coef(l,jj) * mo_coef(k,jj)
e_new += mo_coef(j,ii) * mo_coef(i,ii) * array(j,i,l,k) * mo_coef(l,jj) * mo_coef(k,jj)
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
! if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
! accu_relat += contrib/dabs(array_ref(j,i,l,k))
! endif
enddo
enddo
enddo
enddo
enddo
enddo
print*,'e_ref = ',e_ref
print*,'e_new = ',e_new
! print*,'accu_abs = ',accu_abs/dble(ao_num)**4
! print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_int2_u2_env2
implicit none
integer :: i,j,ipoint,k,l
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += int2_u2_env2_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += int2_u2_env2(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_int2_u2_env2'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_int2_u_grad1u_x_env2
implicit none
integer :: i,j,ipoint,k,l,m
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
do m = 1, 3
array(j,i,l,k) += int2_u_grad1u_x_env2_test(j,i,ipoint,m) * aos_grad_in_r_array_transp(m,k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += int2_u_grad1u_x_env2 (j,i,ipoint,m) * aos_grad_in_r_array_transp(m,k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_int2_u_grad1u_x_env2'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
subroutine routine_v_ij_u_cst_mu_env
implicit none
integer :: i,j,ipoint,k,l
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += v_ij_u_cst_mu_env_test(j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += v_ij_u_cst_mu_env_fit (j,i,ipoint) * aos_in_r_array(k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'routine_v_ij_u_cst_mu_env'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
! ---
subroutine test_fock_3e_uhf_ao()
implicit none
integer :: i, j
double precision :: diff_tot, diff_ij, thr_ih, norm
double precision, allocatable :: fock_3e_uhf_ao_a_mo(:,:), fock_3e_uhf_ao_b_mo(:,:)
thr_ih = 1d-7
PROVIDE fock_a_tot_3e_bi_orth fock_b_tot_3e_bi_orth
PROVIDE fock_3e_uhf_ao_a fock_3e_uhf_ao_b
! ---
allocate(fock_3e_uhf_ao_a_mo(mo_num,mo_num))
call ao_to_mo_bi_ortho( fock_3e_uhf_ao_a , size(fock_3e_uhf_ao_a , 1) &
, fock_3e_uhf_ao_a_mo, size(fock_3e_uhf_ao_a_mo, 1) )
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
diff_ij = dabs(fock_3e_uhf_ao_a_mo(j,i) - fock_a_tot_3e_bi_orth(j,i))
if(diff_ij .gt. thr_ih) then
print *, ' difference on ', j, i
print *, ' MANU : ', fock_a_tot_3e_bi_orth(j,i)
print *, ' UHF : ', fock_3e_uhf_ao_a_mo (j,i)
!stop
endif
norm += dabs(fock_a_tot_3e_bi_orth(j,i))
diff_tot += diff_ij
enddo
enddo
print *, ' diff on F_a = ', diff_tot / norm
print *, ' '
deallocate(fock_3e_uhf_ao_a_mo)
! ---
allocate(fock_3e_uhf_ao_b_mo(mo_num,mo_num))
call ao_to_mo_bi_ortho( fock_3e_uhf_ao_b , size(fock_3e_uhf_ao_b , 1) &
, fock_3e_uhf_ao_b_mo, size(fock_3e_uhf_ao_b_mo, 1) )
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
diff_ij = dabs(fock_3e_uhf_ao_b_mo(j,i) - fock_b_tot_3e_bi_orth(j,i))
if(diff_ij .gt. thr_ih) then
print *, ' difference on ', j, i
print *, ' MANU : ', fock_b_tot_3e_bi_orth(j,i)
print *, ' UHF : ', fock_3e_uhf_ao_b_mo (j,i)
!stop
endif
norm += dabs(fock_b_tot_3e_bi_orth(j,i))
diff_tot += diff_ij
enddo
enddo
print *, ' diff on F_b = ', diff_tot/norm
print *, ' '
deallocate(fock_3e_uhf_ao_b_mo)
! ---
end subroutine test_fock_3e_uhf_ao()
! ---
subroutine test_fock_3e_uhf_mo()
implicit none
integer :: i, j
double precision :: diff_tot, diff_ij, thr_ih, norm
thr_ih = 1d-12
PROVIDE fock_a_tot_3e_bi_orth fock_b_tot_3e_bi_orth
PROVIDE fock_3e_uhf_mo_a fock_3e_uhf_mo_b
! ---
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
diff_ij = dabs(fock_3e_uhf_mo_a(j,i) - fock_a_tot_3e_bi_orth(j,i))
if(diff_ij .gt. thr_ih) then
print *, ' difference on ', j, i
print *, ' MANU : ', fock_a_tot_3e_bi_orth(j,i)
print *, ' UHF : ', fock_3e_uhf_mo_a (j,i)
!stop
endif
norm += dabs(fock_a_tot_3e_bi_orth(j,i))
diff_tot += diff_ij
enddo
enddo
print *, ' diff on F_a = ', diff_tot / norm
print *, ' norm_a = ', norm
print *, ' '
! ---
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
diff_ij = dabs(fock_3e_uhf_mo_b(j,i) - fock_b_tot_3e_bi_orth(j,i))
if(diff_ij .gt. thr_ih) then
print *, ' difference on ', j, i
print *, ' MANU : ', fock_b_tot_3e_bi_orth(j,i)
print *, ' UHF : ', fock_3e_uhf_mo_b (j,i)
!stop
endif
norm += dabs(fock_b_tot_3e_bi_orth(j,i))
diff_tot += diff_ij
enddo
enddo
print *, ' diff on F_b = ', diff_tot/norm
print *, ' norm_b = ', norm
print *, ' '
! ---
end
! ---
subroutine test_grid_points_ao
implicit none
integer :: i,j,ipoint,icount,icount_good, icount_bad,icount_full
double precision :: thr
thr = 1.d-10
! print*,'max_n_pts_grid_ao_prod = ',max_n_pts_grid_ao_prod
! print*,'n_pts_grid_ao_prod'
do i = 1, ao_num
do j = i, ao_num
icount = 0
icount_good = 0
icount_bad = 0
icount_full = 0
do ipoint = 1, n_points_final_grid
! if(dabs(int2_u_grad1u_x_env2_test(j,i,ipoint,1)) &
! + dabs(int2_u_grad1u_x_env2_test(j,i,ipoint,2)) &
! + dabs(int2_u_grad1u_x_env2_test(j,i,ipoint,3)) )
! if(dabs(int2_u2_env2_test(j,i,ipoint)).gt.thr)then
! icount += 1
! endif
if(dabs(v_ij_u_cst_mu_env_ng_1_test(j,i,ipoint)).gt.thr*0.1d0)then
icount_full += 1
endif
if(dabs(v_ij_u_cst_mu_env_test(j,i,ipoint)).gt.thr)then
icount += 1
if(dabs(v_ij_u_cst_mu_env_ng_1_test(j,i,ipoint)).gt.thr*0.1d0)then
icount_good += 1
else
print*,j,i,ipoint
print*,dabs(v_ij_u_cst_mu_env_test(j,i,ipoint)), dabs(v_ij_u_cst_mu_env_ng_1_test(j,i,ipoint)),dabs(v_ij_u_cst_mu_env_ng_1_test(j,i,ipoint))/dabs(v_ij_u_cst_mu_env_test(j,i,ipoint))
icount_bad += 1
endif
endif
! if(dabs(v_ij_u_cst_mu_env_ng_1_test(j,i,ipoint)).gt.thr)then
! endif
enddo
print*,''
print*,j,i
print*,icount,icount_full, icount_bad!,n_pts_grid_ao_prod(j,i)
print*,dble(icount)/dble(n_points_final_grid),dble(icount_full)/dble(n_points_final_grid)
! dble(n_pts_grid_ao_prod(j,i))/dble(n_points_final_grid)
! if(icount.gt.n_pts_grid_ao_prod(j,i))then
! print*,'pb !!'
! endif
enddo
enddo
end
subroutine test_int_gauss
implicit none
integer :: i,j
print*,'center'
do i = 1, ao_num
do j = i, ao_num
print*,j,i
print*,ao_prod_sigma(j,i),ao_overlap_abs_grid(j,i)
print*,ao_prod_center(1:3,j,i)
enddo
enddo
print*,''
double precision :: weight, r(3),integral_1,pi,center(3),f_r,alpha,distance,integral_2
center = 0.d0
pi = dacos(-1.d0)
integral_1 = 0.d0
integral_2 = 0.d0
alpha = 0.75d0
do i = 1, n_points_final_grid
! you get x, y and z of the ith grid point
r(1) = final_grid_points(1,i)
r(2) = final_grid_points(2,i)
r(3) = final_grid_points(3,i)
weight = final_weight_at_r_vector(i)
distance = dsqrt( (r(1) - center(1))**2 + (r(2) - center(2))**2 + (r(3) - center(3))**2 )
f_r = dexp(-alpha * distance*distance)
! you add the contribution of the grid point to the integral
integral_1 += f_r * weight
integral_2 += f_r * distance * weight
enddo
print*,'integral_1 =',integral_1
print*,'(pi/alpha)**1.5 =',(pi / alpha)**1.5
print*,'integral_2 =',integral_2
print*,'(pi/alpha)**1.5 =',2.d0*pi / (alpha)**2
end
! ---
subroutine test_two_e_tc_non_hermit_integral()
implicit none
integer :: i, j
double precision :: diff_tot, diff, thr_ih, norm
thr_ih = 1d-10
PROVIDE two_e_tc_non_hermit_integral_beta two_e_tc_non_hermit_integral_alpha
PROVIDE two_e_tc_non_hermit_integral_seq_beta two_e_tc_non_hermit_integral_seq_alpha
! ---
norm = 0.d0
diff_tot = 0.d0
do i = 1, ao_num
do j = 1, ao_num
diff = dabs(two_e_tc_non_hermit_integral_seq_alpha(j,i) - two_e_tc_non_hermit_integral_alpha(j,i))
if(diff .gt. thr_ih) then
print *, ' difference on ', j, i
print *, ' seq : ', two_e_tc_non_hermit_integral_seq_alpha(j,i)
print *, ' // : ', two_e_tc_non_hermit_integral_alpha (j,i)
!stop
endif
norm += dabs(two_e_tc_non_hermit_integral_seq_alpha(j,i))
diff_tot += diff
enddo
enddo
print *, ' diff tot a = ', diff_tot / norm
print *, ' norm a = ', norm
print *, ' '
! ---
norm = 0.d0
diff_tot = 0.d0
do i = 1, ao_num
do j = 1, ao_num
diff = dabs(two_e_tc_non_hermit_integral_seq_beta(j,i) - two_e_tc_non_hermit_integral_beta(j,i))
if(diff .gt. thr_ih) then
print *, ' difference on ', j, i
print *, ' seq : ', two_e_tc_non_hermit_integral_seq_beta(j,i)
print *, ' // : ', two_e_tc_non_hermit_integral_beta (j,i)
!stop
endif
norm += dabs(two_e_tc_non_hermit_integral_seq_beta(j,i))
diff_tot += diff
enddo
enddo
print *, ' diff tot b = ', diff_tot / norm
print *, ' norm b = ', norm
print *, ' '
! ---
return
end
! ---
subroutine test_int2_grad1_u12_ao_test
implicit none
integer :: i,j,ipoint,m,k,l
double precision :: weight,accu_relat, accu_abs, contrib
double precision, allocatable :: array(:,:,:,:), array_ref(:,:,:,:)
allocate(array(ao_num, ao_num, ao_num, ao_num))
array = 0.d0
allocate(array_ref(ao_num, ao_num, ao_num, ao_num))
array_ref = 0.d0
do m = 1, 3
do ipoint = 1, n_points_final_grid
weight = final_weight_at_r_vector(ipoint)
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
array(j,i,l,k) += int2_grad1_u12_ao_test(j,i,ipoint,m) * aos_grad_in_r_array_transp(m,k,ipoint) * aos_in_r_array(l,ipoint) * weight
array_ref(j,i,l,k) += int2_grad1_u12_ao(j,i,ipoint,m) * aos_grad_in_r_array_transp(m,k,ipoint) * aos_in_r_array(l,ipoint) * weight
enddo
enddo
enddo
enddo
enddo
enddo
accu_relat = 0.d0
accu_abs = 0.d0
do k = 1, ao_num
do l = 1, ao_num
do i = 1, ao_num
do j = 1, ao_num
contrib = dabs(array(j,i,l,k) - array_ref(j,i,l,k))
accu_abs += contrib
if(dabs(array_ref(j,i,l,k)).gt.1.d-10)then
accu_relat += contrib/dabs(array_ref(j,i,l,k))
endif
enddo
enddo
enddo
enddo
print*,'******'
print*,'******'
print*,'test_int2_grad1_u12_ao_test'
print*,'accu_abs = ',accu_abs/dble(ao_num)**4
print*,'accu_relat = ',accu_relat/dble(ao_num)**4
end
! ---
subroutine test_fock_3e_uhf_mo_cs()
implicit none
integer :: i, j
double precision :: I_old, I_new
double precision :: diff_tot, diff, thr_ih, norm
! double precision :: t0, t1
! print*, ' Providing fock_a_tot_3e_bi_orth ...'
! call wall_time(t0)
! PROVIDE fock_a_tot_3e_bi_orth
! call wall_time(t1)
! print*, ' Wall time for fock_a_tot_3e_bi_orth =', t1 - t0
PROVIDE fock_3e_uhf_mo_cs fock_3e_uhf_mo_cs_old
thr_ih = 1d-8
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
I_old = fock_3e_uhf_mo_cs_old(j,i)
I_new = fock_3e_uhf_mo_cs (j,i)
diff = dabs(I_old - I_new)
if(diff .gt. thr_ih) then
print *, ' problem in fock_3e_uhf_mo_cs on ', j, i
print *, ' old value = ', I_old
print *, ' new value = ', I_new
!stop
endif
norm += dabs(I_old)
diff_tot += diff
enddo
enddo
print *, ' diff tot (%) = ', 100.d0 * diff_tot / norm
return
end
! ---
subroutine test_fock_3e_uhf_mo_a()
implicit none
integer :: i, j
double precision :: I_old, I_new
double precision :: diff_tot, diff, thr_ih, norm
PROVIDE fock_3e_uhf_mo_a fock_3e_uhf_mo_a_old
thr_ih = 1d-8
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
I_old = fock_3e_uhf_mo_a_old(j,i)
I_new = fock_3e_uhf_mo_a (j,i)
diff = dabs(I_old - I_new)
if(diff .gt. thr_ih) then
print *, ' problem in fock_3e_uhf_mo_a on ', j, i
print *, ' old value = ', I_old
print *, ' new value = ', I_new
!stop
endif
norm += dabs(I_old)
diff_tot += diff
enddo
enddo
print *, ' diff tot (%) = ', 100.d0 * diff_tot / norm
return
end
! ---
subroutine test_fock_3e_uhf_mo_b()
implicit none
integer :: i, j
double precision :: I_old, I_new
double precision :: diff_tot, diff, thr_ih, norm
PROVIDE fock_3e_uhf_mo_b fock_3e_uhf_mo_b_old
thr_ih = 1d-8
norm = 0.d0
diff_tot = 0.d0
do i = 1, mo_num
do j = 1, mo_num
I_old = fock_3e_uhf_mo_b_old(j,i)
I_new = fock_3e_uhf_mo_b (j,i)
diff = dabs(I_old - I_new)
if(diff .gt. thr_ih) then
print *, ' problem in fock_3e_uhf_mo_b on ', j, i
print *, ' old value = ', I_old
print *, ' new value = ', I_new
!stop
endif
norm += dabs(I_old)
diff_tot += diff
enddo
enddo
print *, ' diff tot (%) = ', 100.d0 * diff_tot / norm
return
end
! ---

View File

@ -1,189 +0,0 @@
subroutine contrib_3e_diag_sss(i, j, k, integral)
BEGIN_DOC
! returns the pure same spin contribution to diagonal matrix element of 3e term
END_DOC
implicit none
integer, intent(in) :: i, j, k
double precision, intent(out) :: integral
double precision :: direct_int, exch_13_int, exch_23_int, exch_12_int, c_3_int, c_minus_3_int
call give_integrals_3_body_bi_ort(i, k, j, i, k, j, direct_int )!!! < i k j | i k j >
call give_integrals_3_body_bi_ort(i, k, j, j, i, k, c_3_int) ! < i k j | j i k >
call give_integrals_3_body_bi_ort(i, k, j, k, j, i, c_minus_3_int)! < i k j | k j i >
integral = direct_int + c_3_int + c_minus_3_int
! negative terms :: exchange contrib
call give_integrals_3_body_bi_ort(i, k, j, j, k, i, exch_13_int)!!! < i k j | j k i > : E_13
call give_integrals_3_body_bi_ort(i, k, j, i, j, k, exch_23_int)!!! < i k j | i j k > : E_23
call give_integrals_3_body_bi_ort(i, k, j, k, i, j, exch_12_int)!!! < i k j | k i j > : E_12
integral += - exch_13_int - exch_23_int - exch_12_int
integral = -integral
end
! ---
subroutine contrib_3e_diag_soo(i,j,k,integral)
implicit none
integer, intent(in) :: i,j,k
BEGIN_DOC
! returns the pure same spin contribution to diagonal matrix element of 3e term
END_DOC
double precision, intent(out) :: integral
double precision :: direct_int, exch_23_int
call give_integrals_3_body_bi_ort(i, k, j, i, k, j, direct_int) ! < i k j | i k j >
call give_integrals_3_body_bi_ort(i, k, j, i, j, k, exch_23_int)! < i k j | i j k > : E_23
integral = direct_int - exch_23_int
integral = -integral
end
subroutine give_aaa_contrib_bis(integral_aaa)
implicit none
double precision, intent(out) :: integral_aaa
double precision :: integral
integer :: i,j,k
integral_aaa = 0.d0
do i = 1, elec_alpha_num
do j = i+1, elec_alpha_num
do k = j+1, elec_alpha_num
call contrib_3e_diag_sss(i,j,k,integral)
integral_aaa += integral
enddo
enddo
enddo
end
! ---
subroutine give_aaa_contrib(integral_aaa)
implicit none
integer :: i, j, k
double precision :: integral
double precision, intent(out) :: integral_aaa
integral_aaa = 0.d0
do i = 1, elec_alpha_num
do j = 1, elec_alpha_num
do k = 1, elec_alpha_num
call contrib_3e_diag_sss(i, j, k, integral)
integral_aaa += integral
enddo
enddo
enddo
integral_aaa *= 1.d0/6.d0
return
end
! ---
subroutine give_aab_contrib(integral_aab)
implicit none
double precision, intent(out) :: integral_aab
double precision :: integral
integer :: i,j,k
integral_aab = 0.d0
do i = 1, elec_beta_num
do j = 1, elec_alpha_num
do k = 1, elec_alpha_num
call contrib_3e_diag_soo(i,j,k,integral)
integral_aab += integral
enddo
enddo
enddo
integral_aab *= 0.5d0
end
subroutine give_aab_contrib_bis(integral_aab)
implicit none
double precision, intent(out) :: integral_aab
double precision :: integral
integer :: i,j,k
integral_aab = 0.d0
do i = 1, elec_beta_num
do j = 1, elec_alpha_num
do k = j+1, elec_alpha_num
call contrib_3e_diag_soo(i,j,k,integral)
integral_aab += integral
enddo
enddo
enddo
end
subroutine give_abb_contrib(integral_abb)
implicit none
double precision, intent(out) :: integral_abb
double precision :: integral
integer :: i,j,k
integral_abb = 0.d0
do i = 1, elec_alpha_num
do j = 1, elec_beta_num
do k = 1, elec_beta_num
call contrib_3e_diag_soo(i,j,k,integral)
integral_abb += integral
enddo
enddo
enddo
integral_abb *= 0.5d0
end
subroutine give_abb_contrib_bis(integral_abb)
implicit none
double precision, intent(out) :: integral_abb
double precision :: integral
integer :: i,j,k
integral_abb = 0.d0
do i = 1, elec_alpha_num
do j = 1, elec_beta_num
do k = j+1, elec_beta_num
call contrib_3e_diag_soo(i,j,k,integral)
integral_abb += integral
enddo
enddo
enddo
end
subroutine give_bbb_contrib_bis(integral_bbb)
implicit none
double precision, intent(out) :: integral_bbb
double precision :: integral
integer :: i,j,k
integral_bbb = 0.d0
do i = 1, elec_beta_num
do j = i+1, elec_beta_num
do k = j+1, elec_beta_num
call contrib_3e_diag_sss(i,j,k,integral)
integral_bbb += integral
enddo
enddo
enddo
end
subroutine give_bbb_contrib(integral_bbb)
implicit none
double precision, intent(out) :: integral_bbb
double precision :: integral
integer :: i,j,k
integral_bbb = 0.d0
do i = 1, elec_beta_num
do j = 1, elec_beta_num
do k = 1, elec_beta_num
call contrib_3e_diag_sss(i,j,k,integral)
integral_bbb += integral
enddo
enddo
enddo
integral_bbb *= 1.d0/6.d0
end

View File

@ -4,11 +4,9 @@ program write_ao_2e_tc_integ
implicit none
PROVIDE j1e_type
PROVIDE j2e_type
print *, ' j1e_type = ', j1e_type
print *, ' j2e_type = ', j2e_type
print *, ' j1e_type = ', j1e_type
print *, ' env_type = ', env_type
my_grid_becke = .True.
PROVIDE tc_grid1_a tc_grid1_r

View File

@ -70,17 +70,6 @@ END_PROVIDER
index_final_points_extra(2,i_count) = i
index_final_points_extra(3,i_count) = j
index_final_points_extra_reverse(k,i,j) = i_count
if(final_weight_at_r_vector_extra(i_count) .lt. 0.d0) then
print *, ' !!! WARNING !!!'
print *, ' negative weight !!!!'
print *, i_count, final_weight_at_r_vector_extra(i_count)
if(dabs(final_weight_at_r_vector_extra(i_count)) .lt. 1d-10) then
final_weight_at_r_vector_extra(i_count) = 0.d0
else
stop
endif
endif
enddo
enddo
enddo

View File

@ -67,17 +67,6 @@ END_PROVIDER
index_final_points(2,i_count) = i
index_final_points(3,i_count) = j
index_final_points_reverse(k,i,j) = i_count
if(final_weight_at_r_vector(i_count) .lt. 0.d0) then
print *, ' !!! WARNING !!!'
print *, ' negative weight !!!!'
print *, i_count, final_weight_at_r_vector(i_count)
if(dabs(final_weight_at_r_vector(i_count)) .lt. 1d-10) then
final_weight_at_r_vector(i_count) = 0.d0
else
stop
endif
endif
enddo
enddo
enddo

View File

@ -319,7 +319,7 @@ call omp_set_max_active_levels(4)
! \end{equation}
! We need a vector to use the gradient. Here the gradient is a
! antisymetric matrix so we can transform it in a vector of length
! antisymmetric matrix so we can transform it in a vector of length
! mo_num*(mo_num-1)/2.
! Here we do these two things at the same time.

View File

@ -284,7 +284,7 @@ call omp_set_max_active_levels(4)
! \end{equation}
! We need a vector to use the gradient. Here the gradient is a
! antisymetric matrix so we can transform it in a vector of length
! antisymmetric matrix so we can transform it in a vector of length
! mo_num*(mo_num-1)/2.
! Here we do these two things at the same time.

View File

@ -576,7 +576,7 @@ logical function is_same_spin(sigma_1, sigma_2)
is_same_spin = .false.
endif
end function is_same_spin
end
! ---
@ -596,7 +596,7 @@ function Kronecker_delta(i, j) result(delta)
delta = 0.d0
endif
end function Kronecker_delta
end
! ---
@ -634,7 +634,81 @@ subroutine diagonalize_sym_matrix(N, A, e)
print*,'Problem in diagonalize_sym_matrix (dsyev)!!'
endif
end subroutine diagonalize_sym_matrix
end
! ---
subroutine give_degen(A, n, shift, list_degen, n_degen_list)
BEGIN_DOC
! returns n_degen_list :: the number of degenerated SET of elements (i.e. with |A(i)-A(i+1)| below shift)
!
! for each of these sets, list_degen(1,i) = first degenerate element of the set i,
!
! list_degen(2,i) = last degenerate element of the set i.
END_DOC
implicit none
double precision, intent(in) :: A(n)
double precision, intent(in) :: shift
integer, intent(in) :: n
integer, intent(out) :: list_degen(2,n), n_degen_list
integer :: i, j, n_degen, k
logical :: keep_on
double precision, allocatable :: Aw(:)
list_degen = -1
allocate(Aw(n))
Aw = A
i=1
k = 0
do while(i.lt.n)
if(dabs(Aw(i)-Aw(i+1)).lt.shift)then
k+=1
j=1
list_degen(1,k) = i
keep_on = .True.
do while(keep_on)
if(i+j.gt.n)then
keep_on = .False.
exit
endif
if(dabs(Aw(i)-Aw(i+j)).lt.shift)then
j+=1
else
keep_on=.False.
exit
endif
enddo
n_degen = j
list_degen(2,k) = list_degen(1,k)-1 + n_degen
j=0
keep_on = .True.
do while(keep_on)
if(i+j+1.gt.n)then
keep_on = .False.
exit
endif
if(dabs(Aw(i+j)-Aw(i+j+1)).lt.shift)then
Aw(i+j) += (j-n_degen/2) * shift
j+=1
else
keep_on = .False.
exit
endif
enddo
Aw(i+n_degen-1) += (n_degen-1-n_degen/2) * shift
i+=n_degen
else
i+=1
endif
enddo
n_degen_list = k
end
! ---