9
1
mirror of https://github.com/QuantumPackage/qp2.git synced 2024-12-21 11:03:29 +01:00

Merge pull request #140 from QuantumPackage/dev

Dev
This commit is contained in:
Emmanuel Giner 2020-11-08 18:29:38 +01:00 committed by GitHub
commit 25b99f1799
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
37 changed files with 272 additions and 2957 deletions

View File

@ -4,8 +4,8 @@ type units =
| Angstrom
;;
let angstrom_to_bohr = 1. /. 0.52917721092
let bohr_to_angstrom = 0.52917721092
let angstrom_to_bohr = 1. /. 0.52917721067121
let bohr_to_angstrom = 0.52917721067121
;;

View File

@ -1862,6 +1862,7 @@ double precision function int_prod_bessel(l,gam,n,m,a,b,arg)
qk = dble(q)
two_qkmp1 = 2.d0*(qk+mk)+1.d0
do k=0,q-1
! possible FPE here. To be checked
s_q_k = two_qkmp1*qk*inverses(k)*s_q_k
! if (s_q_k < 1.d-32) then
! s_q_k = 0.d0

View File

@ -436,7 +436,7 @@ BEGIN_PROVIDER [ double precision, ao_two_e_integral_schwartz,(ao_num,ao_num) ]
!$OMP SCHEDULE(dynamic)
do i=1,ao_num
do k=1,i
ao_two_e_integral_schwartz(i,k) = dsqrt(ao_two_e_integral(i,k,i,k))
ao_two_e_integral_schwartz(i,k) = dsqrt(ao_two_e_integral(i,i,k,k))
ao_two_e_integral_schwartz(k,i) = ao_two_e_integral_schwartz(i,k)
enddo
enddo

View File

@ -1,49 +0,0 @@
#!/usr/bin/env bats
source $QP_ROOT/tests/bats/common.bats.sh
source $QP_ROOT/quantum_package.rc
function run_stoch() {
thresh=$2
test_exe casscf || skip
qp set perturbation do_pt2 True
qp set determinants n_det_max $3
qp set davidson threshold_davidson 1.e-10
qp set davidson n_states_diag 4
qp run casscf | tee casscf.out
energy1="$(ezfio get casscf energy_pt2 | tr '[]' ' ' | cut -d ',' -f 1)"
eq $energy1 $1 $thresh
}
@test "F2" { # 18.0198s
rm -rf f2_casscf
qp_create_ezfio -b aug-cc-pvdz ../input/f2.zmt -o f2_casscf
qp set_file f2_casscf
qp run scf
qp set_mo_class --core="[1-6,8-9]" --act="[7,10]" --virt="[11-46]"
run_stoch -198.773366970 1.e-4 100000
}
@test "N2" { # 18.0198s
rm -rf n2_casscf
qp_create_ezfio -b aug-cc-pvdz ../input/n2.xyz -o n2_casscf
qp set_file n2_casscf
qp run scf
qp set_mo_class --core="[1-4]" --act="[5-10]" --virt="[11-46]"
run_stoch -109.0961643162 1.e-4 100000
}
@test "N2_stretched" {
rm -rf n2_stretched_casscf
qp_create_ezfio -b aug-cc-pvdz -m 7 ../input/n2_stretched.xyz -o n2_stretched_casscf
qp set_file n2_stretched_casscf
qp run scf | tee scf.out
qp set_mo_class --core="[1-4]" --act="[5-10]" --virt="[11-46]"
qp set electrons elec_alpha_num 7
qp set electrons elec_beta_num 7
run_stoch -108.7860471300 1.e-4 100000
#
}

View File

@ -1,31 +0,0 @@
[energy]
type: double precision
doc: Calculated Selected |FCI| energy
interface: ezfio
size: (determinants.n_states)
[energy_pt2]
type: double precision
doc: Calculated |FCI| energy + |PT2|
interface: ezfio
size: (determinants.n_states)
[cisd_guess]
type: logical
doc: If true, the CASSCF starts with a CISD wave function
interface: ezfio,provider,ocaml
default: True
[state_following_casscf]
type: logical
doc: If |true|, the CASSCF will try to follow the guess CI vector and orbitals
interface: ezfio,provider,ocaml
default: False
[level_shift_casscf]
type: Positive_float
doc: Energy shift on the virtual MOs to improve SCF convergence
interface: ezfio,provider,ocaml
default: 0.005

View File

@ -1 +0,0 @@
the CASCF can be obtained if a proper guess is given to the WF part

View File

@ -1,4 +0,0 @@
cipsi
selectors_full
generators_cas
two_body_rdm

View File

@ -1,5 +0,0 @@
======
casscf
======
|CASSCF| program with the CIPSI algorithm.

View File

@ -1,6 +0,0 @@
! -*- F90 -*-
BEGIN_PROVIDER [logical, bavard]
! bavard=.true.
bavard=.false.
END_PROVIDER

View File

@ -1,155 +0,0 @@
BEGIN_PROVIDER [real*8, bielec_PQxx, (mo_num, mo_num,n_core_inact_act_orb,n_core_inact_act_orb)]
BEGIN_DOC
! bielec_PQxx : integral (pq|xx) with p,q arbitrary, x core or active
! indices are unshifted orbital numbers
END_DOC
implicit none
integer :: i,j,ii,jj,p,q,i3,j3,t3,v3
real*8 :: mo_two_e_integral
bielec_PQxx(:,:,:,:) = 0.d0
PROVIDE mo_two_e_integrals_in_map
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP PRIVATE(i,ii,j,jj,i3,j3) &
!$OMP SHARED(n_core_inact_orb,list_core_inact,mo_num,bielec_PQxx, &
!$OMP n_act_orb,mo_integrals_map,list_act)
!$OMP DO
do i=1,n_core_inact_orb
ii=list_core_inact(i)
do j=i,n_core_inact_orb
jj=list_core_inact(j)
call get_mo_two_e_integrals_i1j1(ii,jj,mo_num,bielec_PQxx(1,1,i,j),mo_integrals_map)
bielec_PQxx(:,:,j,i)=bielec_PQxx(:,:,i,j)
end do
do j=1,n_act_orb
jj=list_act(j)
j3=j+n_core_inact_orb
call get_mo_two_e_integrals_i1j1(ii,jj,mo_num,bielec_PQxx(1,1,i,j3),mo_integrals_map)
bielec_PQxx(:,:,j3,i)=bielec_PQxx(:,:,i,j3)
end do
end do
!$OMP END DO
!$OMP DO
do i=1,n_act_orb
ii=list_act(i)
i3=i+n_core_inact_orb
do j=i,n_act_orb
jj=list_act(j)
j3=j+n_core_inact_orb
call get_mo_two_e_integrals_i1j1(ii,jj,mo_num,bielec_PQxx(1,1,i3,j3),mo_integrals_map)
bielec_PQxx(:,:,j3,i3)=bielec_PQxx(:,:,i3,j3)
end do
end do
!$OMP END DO
!$OMP END PARALLEL
END_PROVIDER
BEGIN_PROVIDER [real*8, bielec_PxxQ, (mo_num,n_core_inact_act_orb,n_core_inact_act_orb, mo_num)]
BEGIN_DOC
! bielec_PxxQ : integral (px|xq) with p,q arbitrary, x core or active
! indices are unshifted orbital numbers
END_DOC
implicit none
integer :: i,j,ii,jj,p,q,i3,j3,t3,v3
double precision, allocatable :: integrals_array(:,:)
real*8 :: mo_two_e_integral
PROVIDE mo_two_e_integrals_in_map
bielec_PxxQ = 0.d0
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP PRIVATE(i,ii,j,jj,i3,j3,integrals_array) &
!$OMP SHARED(n_core_inact_orb,list_core_inact,mo_num,bielec_PxxQ, &
!$OMP n_act_orb,mo_integrals_map,list_act)
allocate(integrals_array(mo_num,mo_num))
!$OMP DO
do i=1,n_core_inact_orb
ii=list_core_inact(i)
do j=i,n_core_inact_orb
jj=list_core_inact(j)
call get_mo_two_e_integrals_ij(ii,jj,mo_num,integrals_array,mo_integrals_map)
do q=1,mo_num
do p=1,mo_num
bielec_PxxQ(p,i,j,q)=integrals_array(p,q)
bielec_PxxQ(p,j,i,q)=integrals_array(q,p)
end do
end do
end do
do j=1,n_act_orb
jj=list_act(j)
j3=j+n_core_inact_orb
call get_mo_two_e_integrals_ij(ii,jj,mo_num,integrals_array,mo_integrals_map)
do q=1,mo_num
do p=1,mo_num
bielec_PxxQ(p,i,j3,q)=integrals_array(p,q)
bielec_PxxQ(p,j3,i,q)=integrals_array(q,p)
end do
end do
end do
end do
!$OMP END DO
! (ip|qj)
!$OMP DO
do i=1,n_act_orb
ii=list_act(i)
i3=i+n_core_inact_orb
do j=i,n_act_orb
jj=list_act(j)
j3=j+n_core_inact_orb
call get_mo_two_e_integrals_ij(ii,jj,mo_num,integrals_array,mo_integrals_map)
do q=1,mo_num
do p=1,mo_num
bielec_PxxQ(p,i3,j3,q)=integrals_array(p,q)
bielec_PxxQ(p,j3,i3,q)=integrals_array(q,p)
end do
end do
end do
end do
!$OMP END DO
deallocate(integrals_array)
!$OMP END PARALLEL
END_PROVIDER
BEGIN_PROVIDER [real*8, bielecCI, (n_act_orb,n_act_orb,n_act_orb, mo_num)]
BEGIN_DOC
! bielecCI : integrals (tu|vp) with p arbitrary, tuv active
! index p runs over the whole basis, t,u,v only over the active orbitals
END_DOC
implicit none
integer :: i,j,k,p,t,u,v
double precision, external :: mo_two_e_integral
PROVIDE mo_two_e_integrals_in_map
!$OMP PARALLEL DO DEFAULT(NONE) &
!$OMP PRIVATE(i,j,k,p,t,u,v) &
!$OMP SHARED(mo_num,n_act_orb,list_act,bielecCI)
do p=1,mo_num
do j=1,n_act_orb
u=list_act(j)
do k=1,n_act_orb
v=list_act(k)
do i=1,n_act_orb
t=list_act(i)
bielecCI(i,k,j,p) = mo_two_e_integral(t,u,v,p)
end do
end do
end do
end do
!$OMP END PARALLEL DO
END_PROVIDER

View File

@ -1,369 +0,0 @@
BEGIN_PROVIDER [real*8, bielec_PQxx_no, (mo_num, mo_num,n_core_inact_act_orb,n_core_inact_act_orb)]
BEGIN_DOC
! integral (pq|xx) in the basis of natural MOs
! indices are unshifted orbital numbers
END_DOC
implicit none
integer :: i,j,k,l,t,u,p,q
double precision, allocatable :: f(:,:,:), d(:,:,:)
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP PRIVATE(j,k,l,p,d,f) &
!$OMP SHARED(n_core_inact_act_orb,mo_num,n_act_orb,n_core_inact_orb, &
!$OMP bielec_PQxx_no,bielec_PQxx,list_act,natorbsCI)
allocate (f(n_act_orb,mo_num,n_core_inact_act_orb), &
d(n_act_orb,mo_num,n_core_inact_act_orb))
!$OMP DO
do l=1,n_core_inact_act_orb
bielec_PQxx_no(:,:,:,l) = bielec_PQxx(:,:,:,l)
do k=1,n_core_inact_act_orb
do j=1,mo_num
do p=1,n_act_orb
f(p,j,k)=bielec_PQxx_no(list_act(p),j,k,l)
end do
end do
end do
call dgemm('T','N',n_act_orb,mo_num*n_core_inact_act_orb,n_act_orb,1.d0, &
natorbsCI, size(natorbsCI,1), &
f, n_act_orb, &
0.d0, &
d, n_act_orb)
do k=1,n_core_inact_act_orb
do j=1,mo_num
do p=1,n_act_orb
bielec_PQxx_no(list_act(p),j,k,l)=d(p,j,k)
end do
end do
do j=1,mo_num
do p=1,n_act_orb
f(p,j,k)=bielec_PQxx_no(j,list_act(p),k,l)
end do
end do
end do
call dgemm('T','N',n_act_orb,mo_num*n_core_inact_act_orb,n_act_orb,1.d0, &
natorbsCI, n_act_orb, &
f, n_act_orb, &
0.d0, &
d, n_act_orb)
do k=1,n_core_inact_act_orb
do p=1,n_act_orb
do j=1,mo_num
bielec_PQxx_no(j,list_act(p),k,l)=d(p,j,k)
end do
end do
end do
end do
!$OMP END DO NOWAIT
deallocate (f,d)
allocate (f(mo_num,mo_num,n_act_orb),d(mo_num,mo_num,n_act_orb))
!$OMP DO
do l=1,n_core_inact_act_orb
do p=1,n_act_orb
do k=1,mo_num
do j=1,mo_num
f(j,k,p) = bielec_PQxx_no(j,k,n_core_inact_orb+p,l)
end do
end do
end do
call dgemm('N','N',mo_num*mo_num,n_act_orb,n_act_orb,1.d0, &
f, mo_num*mo_num, &
natorbsCI, n_act_orb, &
0.d0, &
d, mo_num*mo_num)
do p=1,n_act_orb
do k=1,mo_num
do j=1,mo_num
bielec_PQxx_no(j,k,n_core_inact_orb+p,l)=d(j,k,p)
end do
end do
end do
end do
!$OMP END DO NOWAIT
!$OMP BARRIER
!$OMP DO
do l=1,n_core_inact_act_orb
do p=1,n_act_orb
do k=1,mo_num
do j=1,mo_num
f(j,k,p) = bielec_PQxx_no(j,k,l,n_core_inact_orb+p)
end do
end do
end do
call dgemm('N','N',mo_num*mo_num,n_act_orb,n_act_orb,1.d0, &
f, mo_num*mo_num, &
natorbsCI, n_act_orb, &
0.d0, &
d, mo_num*mo_num)
do p=1,n_act_orb
do k=1,mo_num
do j=1,mo_num
bielec_PQxx_no(j,k,l,n_core_inact_orb+p)=d(j,k,p)
end do
end do
end do
end do
!$OMP END DO
deallocate (f,d)
!$OMP END PARALLEL
END_PROVIDER
BEGIN_PROVIDER [real*8, bielec_PxxQ_no, (mo_num,n_core_inact_act_orb,n_core_inact_act_orb, mo_num)]
BEGIN_DOC
! integral (px|xq) in the basis of natural MOs
! indices are unshifted orbital numbers
END_DOC
implicit none
integer :: i,j,k,l,t,u,p,q
double precision, allocatable :: f(:,:,:), d(:,:,:)
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP PRIVATE(j,k,l,p,d,f) &
!$OMP SHARED(n_core_inact_act_orb,mo_num,n_act_orb,n_core_inact_orb, &
!$OMP bielec_PxxQ_no,bielec_PxxQ,list_act,natorbsCI)
allocate (f(n_act_orb,n_core_inact_act_orb,n_core_inact_act_orb), &
d(n_act_orb,n_core_inact_act_orb,n_core_inact_act_orb))
!$OMP DO
do j=1,mo_num
bielec_PxxQ_no(:,:,:,j) = bielec_PxxQ(:,:,:,j)
do l=1,n_core_inact_act_orb
do k=1,n_core_inact_act_orb
do p=1,n_act_orb
f(p,k,l) = bielec_PxxQ_no(list_act(p),k,l,j)
end do
end do
end do
call dgemm('T','N',n_act_orb,n_core_inact_act_orb**2,n_act_orb,1.d0, &
natorbsCI, size(natorbsCI,1), &
f, n_act_orb, &
0.d0, &
d, n_act_orb)
do l=1,n_core_inact_act_orb
do k=1,n_core_inact_act_orb
do p=1,n_act_orb
bielec_PxxQ_no(list_act(p),k,l,j)=d(p,k,l)
end do
end do
end do
end do
!$OMP END DO NOWAIT
deallocate (f,d)
allocate (f(n_act_orb,mo_num,n_core_inact_act_orb), &
d(n_act_orb,mo_num,n_core_inact_act_orb))
!$OMP DO
do k=1,mo_num
do l=1,n_core_inact_act_orb
do j=1,mo_num
do p=1,n_act_orb
f(p,j,l) = bielec_PxxQ_no(j,n_core_inact_orb+p,l,k)
end do
end do
end do
call dgemm('T','N',n_act_orb,mo_num*n_core_inact_act_orb,n_act_orb,1.d0, &
natorbsCI, size(natorbsCI,1), &
f, n_act_orb, &
0.d0, &
d, n_act_orb)
do l=1,n_core_inact_act_orb
do j=1,mo_num
do p=1,n_act_orb
bielec_PxxQ_no(j,n_core_inact_orb+p,l,k)=d(p,j,l)
end do
end do
end do
end do
!$OMP END DO NOWAIT
deallocate(f,d)
allocate(f(mo_num,n_core_inact_act_orb,n_act_orb), &
d(mo_num,n_core_inact_act_orb,n_act_orb) )
!$OMP DO
do k=1,mo_num
do p=1,n_act_orb
do l=1,n_core_inact_act_orb
do j=1,mo_num
f(j,l,p) = bielec_PxxQ_no(j,l,n_core_inact_orb+p,k)
end do
end do
end do
call dgemm('N','N',mo_num*n_core_inact_act_orb,n_act_orb,n_act_orb,1.d0, &
f, mo_num*n_core_inact_act_orb, &
natorbsCI, size(natorbsCI,1), &
0.d0, &
d, mo_num*n_core_inact_act_orb)
do p=1,n_act_orb
do l=1,n_core_inact_act_orb
do j=1,mo_num
bielec_PxxQ_no(j,l,n_core_inact_orb+p,k)=d(j,l,p)
end do
end do
end do
end do
!$OMP END DO NOWAIT
!$OMP BARRIER
!$OMP DO
do l=1,n_core_inact_act_orb
do p=1,n_act_orb
do k=1,n_core_inact_act_orb
do j=1,mo_num
f(j,k,p) = bielec_PxxQ_no(j,k,l,n_core_inact_orb+p)
end do
end do
end do
call dgemm('N','N',mo_num*n_core_inact_act_orb,n_act_orb,n_act_orb,1.d0, &
f, mo_num*n_core_inact_act_orb, &
natorbsCI, size(natorbsCI,1), &
0.d0, &
d, mo_num*n_core_inact_act_orb)
do p=1,n_act_orb
do k=1,n_core_inact_act_orb
do j=1,mo_num
bielec_PxxQ_no(j,k,l,n_core_inact_orb+p)=d(j,k,p)
end do
end do
end do
end do
!$OMP END DO NOWAIT
deallocate(f,d)
!$OMP END PARALLEL
END_PROVIDER
BEGIN_PROVIDER [real*8, bielecCI_no, (n_act_orb,n_act_orb,n_act_orb, mo_num)]
BEGIN_DOC
! integrals (tu|vp) in the basis of natural MOs
! index p runs over the whole basis, t,u,v only over the active orbitals
END_DOC
implicit none
integer :: i,j,k,l,t,u,p,q
double precision, allocatable :: f(:,:,:), d(:,:,:)
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP PRIVATE(j,k,l,p,d,f) &
!$OMP SHARED(n_core_inact_act_orb,mo_num,n_act_orb,n_core_inact_orb, &
!$OMP bielecCI_no,bielecCI,list_act,natorbsCI)
allocate (f(n_act_orb,n_act_orb,mo_num), &
d(n_act_orb,n_act_orb,mo_num))
!$OMP DO
do l=1,mo_num
bielecCI_no(:,:,:,l) = bielecCI(:,:,:,l)
do k=1,n_act_orb
do j=1,n_act_orb
do p=1,n_act_orb
f(p,j,k)=bielecCI_no(p,j,k,l)
end do
end do
end do
call dgemm('T','N',n_act_orb,n_act_orb*n_act_orb,n_act_orb,1.d0, &
natorbsCI, size(natorbsCI,1), &
f, n_act_orb, &
0.d0, &
d, n_act_orb)
do k=1,n_act_orb
do j=1,n_act_orb
do p=1,n_act_orb
bielecCI_no(p,j,k,l)=d(p,j,k)
end do
end do
do j=1,n_act_orb
do p=1,n_act_orb
f(p,j,k)=bielecCI_no(j,p,k,l)
end do
end do
end do
call dgemm('T','N',n_act_orb,n_act_orb*n_act_orb,n_act_orb,1.d0, &
natorbsCI, n_act_orb, &
f, n_act_orb, &
0.d0, &
d, n_act_orb)
do k=1,n_act_orb
do p=1,n_act_orb
do j=1,n_act_orb
bielecCI_no(j,p,k,l)=d(p,j,k)
end do
end do
end do
do p=1,n_act_orb
do k=1,n_act_orb
do j=1,n_act_orb
f(j,k,p)=bielecCI_no(j,k,p,l)
end do
end do
end do
call dgemm('N','N',n_act_orb*n_act_orb,n_act_orb,n_act_orb,1.d0, &
f, n_act_orb*n_act_orb, &
natorbsCI, n_act_orb, &
0.d0, &
d, n_act_orb*n_act_orb)
do p=1,n_act_orb
do k=1,n_act_orb
do j=1,n_act_orb
bielecCI_no(j,k,p,l)=d(j,k,p)
end do
end do
end do
end do
!$OMP END DO
!$OMP DO
do l=1,n_act_orb
do p=1,n_act_orb
do k=1,n_act_orb
do j=1,n_act_orb
f(j,k,p)=bielecCI_no(j,k,l,list_act(p))
end do
end do
end do
call dgemm('N','N',n_act_orb*n_act_orb,n_act_orb,n_act_orb,1.d0, &
f, n_act_orb*n_act_orb, &
natorbsCI, n_act_orb, &
0.d0, &
d, n_act_orb*n_act_orb)
do p=1,n_act_orb
do k=1,n_act_orb
do j=1,n_act_orb
bielecCI_no(j,k,l,list_act(p))=d(j,k,p)
end do
end do
end do
end do
!$OMP END DO
deallocate(d,f)
!$OMP END PARALLEL
END_PROVIDER

View File

@ -1,58 +0,0 @@
program casscf
implicit none
BEGIN_DOC
! TODO : Put the documentation of the program here
END_DOC
call reorder_orbitals_for_casscf
no_vvvv_integrals = .True.
touch no_vvvv_integrals
pt2_max = 0.02
SOFT_TOUCH no_vvvv_integrals pt2_max
call run_stochastic_cipsi
call run
end
subroutine run
implicit none
double precision :: energy_old, energy
logical :: converged,state_following_casscf_save
integer :: iteration
converged = .False.
energy = 0.d0
mo_label = "MCSCF"
iteration = 1
state_following_casscf_save = state_following_casscf
state_following_casscf = .True.
touch state_following_casscf
do while (.not.converged)
call run_stochastic_cipsi
energy_old = energy
energy = eone+etwo+ecore
call write_time(6)
call write_int(6,iteration,'CAS-SCF iteration')
call write_double(6,energy,'CAS-SCF energy')
call write_double(6,energy_improvement, 'Predicted energy improvement')
converged = dabs(energy_improvement) < thresh_scf
pt2_max = dabs(energy_improvement / pt2_relative_error)
mo_coef = NewOrbs
mo_occ = occnum
call save_mos
iteration += 1
N_det = max(N_det/2 ,N_states)
psi_det = psi_det_sorted
psi_coef = psi_coef_sorted
read_wf = .True.
call clear_mo_map
SOFT_TOUCH mo_coef N_det pt2_max psi_det psi_coef
if(iteration .gt. 3)then
state_following_casscf = state_following_casscf_save
touch state_following_casscf
endif
enddo
end

View File

@ -1,12 +0,0 @@
BEGIN_PROVIDER [ logical, do_only_1h1p ]
&BEGIN_PROVIDER [ logical, do_only_cas ]
&BEGIN_PROVIDER [ logical, do_ddci ]
implicit none
BEGIN_DOC
! In the CAS case, all those are always false except do_only_cas
END_DOC
do_only_cas = .True.
do_only_1h1p = .False.
do_ddci = .False.
END_PROVIDER

View File

@ -1,63 +0,0 @@
use bitmasks
BEGIN_PROVIDER [real*8, D0tu, (n_act_orb,n_act_orb) ]
implicit none
BEGIN_DOC
! the first-order density matrix in the basis of the starting MOs.
! matrix is state averaged.
END_DOC
integer :: t,u
do u=1,n_act_orb
do t=1,n_act_orb
D0tu(t,u) = one_e_dm_mo_alpha_average( list_act(t), list_act(u) ) + &
one_e_dm_mo_beta_average ( list_act(t), list_act(u) )
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [real*8, P0tuvx, (n_act_orb,n_act_orb,n_act_orb,n_act_orb) ]
BEGIN_DOC
! The second-order density matrix in the basis of the starting MOs ONLY IN THE RANGE OF ACTIVE MOS
! The values are state averaged
!
! We use the spin-free generators of mono-excitations
! E_pq destroys q and creates p
! D_pq = <0|E_pq|0> = D_qp
! P_pqrs = 1/2 <0|E_pq E_rs - delta_qr E_ps|0>
!
! P0tuvx(p,q,r,s) = chemist notation : 1/2 <0|E_pq E_rs - delta_qr E_ps|0>
END_DOC
implicit none
integer :: t,u,v,x
integer :: tt,uu,vv,xx
integer :: mu,nu,istate,ispin,jspin,ihole,ipart,jhole,jpart
integer :: ierr
real*8 :: phase1,phase11,phase12,phase2,phase21,phase22
integer :: nu1,nu2,nu11,nu12,nu21,nu22
integer :: ierr1,ierr2,ierr11,ierr12,ierr21,ierr22
real*8 :: cI_mu(N_states),term
integer(bit_kind), dimension(N_int,2) :: det_mu, det_mu_ex
integer(bit_kind), dimension(N_int,2) :: det_mu_ex1, det_mu_ex11, det_mu_ex12
integer(bit_kind), dimension(N_int,2) :: det_mu_ex2, det_mu_ex21, det_mu_ex22
if (bavard) then
write(6,*) ' providing the 2 body RDM on the active part'
endif
P0tuvx= 0.d0
do istate=1,N_states
do x = 1, n_act_orb
do v = 1, n_act_orb
do u = 1, n_act_orb
do t = 1, n_act_orb
! 1 1 2 2 1 2 1 2
P0tuvx(t,u,v,x) = state_av_act_2_rdm_spin_trace_mo(t,v,u,x)
enddo
enddo
enddo
enddo
enddo
END_PROVIDER

View File

@ -1,125 +0,0 @@
use bitmasks
subroutine do_signed_mono_excitation(key1,key2,nu,ihole,ipart, &
ispin,phase,ierr)
BEGIN_DOC
! we create the mono-excitation, and determine, if possible,
! the phase and the number in the list of determinants
END_DOC
implicit none
integer(bit_kind) :: key1(N_int,2),key2(N_int,2)
integer(bit_kind), allocatable :: keytmp(:,:)
integer :: exc(0:2,2,2),ihole,ipart,ierr,nu,ispin
real*8 :: phase
logical :: found
allocate(keytmp(N_int,2))
nu=-1
phase=1.D0
ierr=0
call det_copy(key1,key2,N_int)
! write(6,*) ' key2 before excitation ',ihole,' -> ',ipart,' spin = ',ispin
! call print_det(key2,N_int)
call do_single_excitation(key2,ihole,ipart,ispin,ierr)
! write(6,*) ' key2 after ',ihole,' -> ',ipart,' spin = ',ispin
! call print_det(key2,N_int)
! write(6,*) ' excitation ',ihole,' -> ',ipart,' gives ierr = ',ierr
if (ierr.eq.1) then
! excitation is possible
! get the phase
call get_single_excitation(key1,key2,exc,phase,N_int)
! get the number in the list
found=.false.
nu=0
!TODO BOTTLENECK
do while (.not.found)
nu+=1
if (nu.gt.N_det) then
! the determinant is possible, but not in the list
found=.true.
nu=-1
else
call det_extract(keytmp,nu,N_int)
integer :: i,ii
found=.true.
do ii=1,2
do i=1,N_int
if (keytmp(i,ii).ne.key2(i,ii)) then
found=.false.
end if
end do
end do
end if
end do
end if
!
! we found the new string, the phase, and possibly the number in the list
!
end subroutine do_signed_mono_excitation
subroutine det_extract(key,nu,Nint)
BEGIN_DOC
! extract a determinant from the list of determinants
END_DOC
implicit none
integer :: ispin,i,nu,Nint
integer(bit_kind) :: key(Nint,2)
do ispin=1,2
do i=1,Nint
key(i,ispin)=psi_det(i,ispin,nu)
end do
end do
end subroutine det_extract
subroutine det_copy(key1,key2,Nint)
use bitmasks ! you need to include the bitmasks_module.f90 features
BEGIN_DOC
! copy a determinant from key1 to key2
END_DOC
implicit none
integer :: ispin,i,Nint
integer(bit_kind) :: key1(Nint,2),key2(Nint,2)
do ispin=1,2
do i=1,Nint
key2(i,ispin)=key1(i,ispin)
end do
end do
end subroutine det_copy
subroutine do_spinfree_mono_excitation(key_in,key_out1,key_out2 &
,nu1,nu2,ihole,ipart,phase1,phase2,ierr,jerr)
BEGIN_DOC
! we create the spin-free mono-excitation E_pq=(a^+_p a_q + a^+_P a_Q)
! we may create two determinants as result
!
END_DOC
implicit none
integer(bit_kind) :: key_in(N_int,2),key_out1(N_int,2)
integer(bit_kind) :: key_out2(N_int,2)
integer :: ihole,ipart,ierr,jerr,nu1,nu2
integer :: ispin
real*8 :: phase1,phase2
! write(6,*) ' applying E_',ipart,ihole,' on determinant '
! call print_det(key_in,N_int)
! spin alpha
ispin=1
call do_signed_mono_excitation(key_in,key_out1,nu1,ihole &
,ipart,ispin,phase1,ierr)
! if (ierr.eq.1) then
! write(6,*) ' 1 result is ',nu1,phase1
! call print_det(key_out1,N_int)
! end if
! spin beta
ispin=2
call do_signed_mono_excitation(key_in,key_out2,nu2,ihole &
,ipart,ispin,phase2,jerr)
! if (jerr.eq.1) then
! write(6,*) ' 2 result is ',nu2,phase2
! call print_det(key_out2,N_int)
! end if
end subroutine do_spinfree_mono_excitation

View File

@ -1,3 +0,0 @@
subroutine driver_optorb
implicit none
end

View File

@ -1,51 +0,0 @@
program print_2rdm
implicit none
BEGIN_DOC
! get the active part of the bielectronic energy on a given wave function.
!
! useful to test the active part of the spin trace 2 rdms
END_DOC
!no_vvvv_integrals = .True.
read_wf = .True.
!touch read_wf no_vvvv_integrals
!call routine
!call routine_bis
call print_grad
end
subroutine print_grad
implicit none
integer :: i
do i = 1, nMonoEx
if(dabs(gradvec2(i)).gt.1.d-5)then
print*,''
print*,i,gradvec2(i),excit(:,i)
endif
enddo
end
subroutine routine
integer :: i,j,k,l
integer :: ii,jj,kk,ll
double precision :: accu(4),twodm,thr,act_twodm2,integral,get_two_e_integral
thr = 1.d-10
accu = 0.d0
do ll = 1, n_act_orb
l = list_act(ll)
do kk = 1, n_act_orb
k = list_act(kk)
do jj = 1, n_act_orb
j = list_act(jj)
do ii = 1, n_act_orb
i = list_act(ii)
integral = get_two_e_integral(i,j,k,l,mo_integrals_map)
accu(1) += state_av_act_2_rdm_spin_trace_mo(ii,jj,kk,ll) * integral
enddo
enddo
enddo
enddo
print*,'accu = ',accu(1)
end

View File

@ -1,74 +0,0 @@
BEGIN_PROVIDER [real*8, gradvec_old, (nMonoEx)]
BEGIN_DOC
! calculate the orbital gradient <Psi| H E_pq |Psi> by hand, i.e. for
! each determinant I we determine the string E_pq |I> (alpha and beta
! separately) and generate <Psi|H E_pq |I>
! sum_I c_I <Psi|H E_pq |I> is then the pq component of the orbital
! gradient
! E_pq = a^+_pa_q + a^+_Pa_Q
END_DOC
implicit none
integer :: ii,tt,aa,indx,ihole,ipart,istate
real*8 :: res
do indx=1,nMonoEx
ihole=excit(1,indx)
ipart=excit(2,indx)
call calc_grad_elem(ihole,ipart,res)
gradvec_old(indx)=res
end do
real*8 :: norm_grad
norm_grad=0.d0
do indx=1,nMonoEx
norm_grad+=gradvec_old(indx)*gradvec_old(indx)
end do
norm_grad=sqrt(norm_grad)
if (bavard) then
write(6,*)
write(6,*) ' Norm of the orbital gradient (via <0|EH|0>) : ', norm_grad
write(6,*)
endif
END_PROVIDER
subroutine calc_grad_elem(ihole,ipart,res)
BEGIN_DOC
! eq 18 of Siegbahn et al, Physica Scripta 1980
! we calculate 2 <Psi| H E_pq | Psi>, q=hole, p=particle
END_DOC
implicit none
integer :: ihole,ipart,mu,iii,ispin,ierr,nu,istate
real*8 :: res
integer(bit_kind), allocatable :: det_mu(:,:),det_mu_ex(:,:)
real*8 :: i_H_psi_array(N_states),phase
allocate(det_mu(N_int,2))
allocate(det_mu_ex(N_int,2))
res=0.D0
do mu=1,n_det
! get the string of the determinant
call det_extract(det_mu,mu,N_int)
do ispin=1,2
! do the monoexcitation on it
call det_copy(det_mu,det_mu_ex,N_int)
call do_signed_mono_excitation(det_mu,det_mu_ex,nu &
,ihole,ipart,ispin,phase,ierr)
if (ierr.eq.1) then
call i_H_psi(det_mu_ex,psi_det,psi_coef,N_int &
,N_det,N_det,N_states,i_H_psi_array)
do istate=1,N_states
res+=i_H_psi_array(istate)*psi_coef(mu,istate)*phase
end do
end if
end do
end do
! state-averaged gradient
res*=2.D0/dble(N_states)
end subroutine calc_grad_elem

View File

@ -1,171 +0,0 @@
use bitmasks
BEGIN_PROVIDER [ integer, nMonoEx ]
BEGIN_DOC
! Number of single excitations
END_DOC
implicit none
nMonoEx=n_core_inact_orb*n_act_orb+n_core_inact_orb*n_virt_orb+n_act_orb*n_virt_orb
END_PROVIDER
BEGIN_PROVIDER [integer, excit, (2,nMonoEx)]
&BEGIN_PROVIDER [character*3, excit_class, (nMonoEx)]
BEGIN_DOC
! a list of the orbitals involved in the excitation
END_DOC
implicit none
integer :: i,t,a,ii,tt,aa,indx
indx=0
do ii=1,n_core_inact_orb
i=list_core_inact(ii)
do tt=1,n_act_orb
t=list_act(tt)
indx+=1
excit(1,indx)=i
excit(2,indx)=t
excit_class(indx)='c-a'
end do
end do
do ii=1,n_core_inact_orb
i=list_core_inact(ii)
do aa=1,n_virt_orb
a=list_virt(aa)
indx+=1
excit(1,indx)=i
excit(2,indx)=a
excit_class(indx)='c-v'
end do
end do
do tt=1,n_act_orb
t=list_act(tt)
do aa=1,n_virt_orb
a=list_virt(aa)
indx+=1
excit(1,indx)=t
excit(2,indx)=a
excit_class(indx)='a-v'
end do
end do
if (bavard) then
write(6,*) ' Filled the table of the Monoexcitations '
do indx=1,nMonoEx
write(6,*) ' ex ',indx,' : ',excit(1,indx),' -> ' &
,excit(2,indx),' ',excit_class(indx)
end do
end if
END_PROVIDER
BEGIN_PROVIDER [real*8, gradvec2, (nMonoEx)]
BEGIN_DOC
! calculate the orbital gradient <Psi| H E_pq |Psi> from density
! matrices and integrals; Siegbahn et al, Phys Scr 1980
! eqs 14 a,b,c
END_DOC
implicit none
integer :: i,t,a,indx
real*8 :: gradvec_it,gradvec_ia,gradvec_ta
real*8 :: norm_grad
indx=0
do i=1,n_core_inact_orb
do t=1,n_act_orb
indx+=1
gradvec2(indx)=gradvec_it(i,t)
end do
end do
do i=1,n_core_inact_orb
do a=1,n_virt_orb
indx+=1
gradvec2(indx)=gradvec_ia(i,a)
end do
end do
do t=1,n_act_orb
do a=1,n_virt_orb
indx+=1
gradvec2(indx)=gradvec_ta(t,a)
end do
end do
norm_grad=0.d0
do indx=1,nMonoEx
norm_grad+=gradvec2(indx)*gradvec2(indx)
end do
norm_grad=sqrt(norm_grad)
write(6,*)
write(6,*) ' Norm of the orbital gradient (via D, P and integrals): ', norm_grad
write(6,*)
END_PROVIDER
real*8 function gradvec_it(i,t)
BEGIN_DOC
! the orbital gradient core/inactive -> active
! we assume natural orbitals
END_DOC
implicit none
integer :: i,t
integer :: ii,tt,v,vv,x,y
integer :: x3,y3
ii=list_core_inact(i)
tt=list_act(t)
gradvec_it=2.D0*(Fipq(tt,ii)+Fapq(tt,ii))
gradvec_it-=occnum(tt)*Fipq(ii,tt)
do v=1,n_act_orb
vv=list_act(v)
do x=1,n_act_orb
x3=x+n_core_inact_orb
do y=1,n_act_orb
y3=y+n_core_inact_orb
gradvec_it-=2.D0*P0tuvx_no(t,v,x,y)*bielec_PQxx_no(ii,vv,x3,y3)
end do
end do
end do
gradvec_it*=2.D0
end function gradvec_it
real*8 function gradvec_ia(i,a)
BEGIN_DOC
! the orbital gradient core/inactive -> virtual
END_DOC
implicit none
integer :: i,a,ii,aa
ii=list_core_inact(i)
aa=list_virt(a)
gradvec_ia=2.D0*(Fipq(aa,ii)+Fapq(aa,ii))
gradvec_ia*=2.D0
end function gradvec_ia
real*8 function gradvec_ta(t,a)
BEGIN_DOC
! the orbital gradient active -> virtual
! we assume natural orbitals
END_DOC
implicit none
integer :: t,a,tt,aa,v,vv,x,y
tt=list_act(t)
aa=list_virt(a)
gradvec_ta=0.D0
gradvec_ta+=occnum(tt)*Fipq(aa,tt)
do v=1,n_act_orb
do x=1,n_act_orb
do y=1,n_act_orb
gradvec_ta+=2.D0*P0tuvx_no(t,v,x,y)*bielecCI_no(x,y,v,aa)
end do
end do
end do
gradvec_ta*=2.D0
end function gradvec_ta

View File

@ -1,656 +0,0 @@
use bitmasks
BEGIN_PROVIDER [real*8, hessmat, (nMonoEx,nMonoEx)]
BEGIN_DOC
! calculate the orbital hessian 2 <Psi| E_pq H E_rs |Psi>
! + <Psi| E_pq E_rs H |Psi> + <Psi| E_rs E_pq H |Psi> by hand,
! determinant per determinant, as for the gradient
!
! we assume that we have natural active orbitals
END_DOC
implicit none
integer :: indx,ihole,ipart
integer :: jndx,jhole,jpart
character*3 :: iexc,jexc
real*8 :: res
if (bavard) then
write(6,*) ' providing Hessian matrix hessmat '
write(6,*) ' nMonoEx = ',nMonoEx
endif
do indx=1,nMonoEx
do jndx=1,nMonoEx
hessmat(indx,jndx)=0.D0
end do
end do
do indx=1,nMonoEx
ihole=excit(1,indx)
ipart=excit(2,indx)
iexc=excit_class(indx)
do jndx=indx,nMonoEx
jhole=excit(1,jndx)
jpart=excit(2,jndx)
jexc=excit_class(jndx)
call calc_hess_elem(ihole,ipart,jhole,jpart,res)
hessmat(indx,jndx)=res
hessmat(jndx,indx)=res
end do
end do
END_PROVIDER
subroutine calc_hess_elem(ihole,ipart,jhole,jpart,res)
BEGIN_DOC
! eq 19 of Siegbahn et al, Physica Scripta 1980
! we calculate 2 <Psi| E_pq H E_rs |Psi>
! + <Psi| E_pq E_rs H |Psi> + <Psi| E_rs E_pq H |Psi>
! average over all states is performed.
! no transition between states.
END_DOC
implicit none
integer :: ihole,ipart,ispin,mu,istate
integer :: jhole,jpart,jspin
integer :: mu_pq, mu_pqrs, mu_rs, mu_rspq, nu_rs,nu
real*8 :: res
integer(bit_kind), allocatable :: det_mu(:,:)
integer(bit_kind), allocatable :: det_nu(:,:)
integer(bit_kind), allocatable :: det_mu_pq(:,:)
integer(bit_kind), allocatable :: det_mu_rs(:,:)
integer(bit_kind), allocatable :: det_nu_rs(:,:)
integer(bit_kind), allocatable :: det_mu_pqrs(:,:)
integer(bit_kind), allocatable :: det_mu_rspq(:,:)
real*8 :: i_H_psi_array(N_states),phase,phase2,phase3
real*8 :: i_H_j_element
allocate(det_mu(N_int,2))
allocate(det_nu(N_int,2))
allocate(det_mu_pq(N_int,2))
allocate(det_mu_rs(N_int,2))
allocate(det_nu_rs(N_int,2))
allocate(det_mu_pqrs(N_int,2))
allocate(det_mu_rspq(N_int,2))
integer :: mu_pq_possible
integer :: mu_rs_possible
integer :: nu_rs_possible
integer :: mu_pqrs_possible
integer :: mu_rspq_possible
res=0.D0
! the terms <0|E E H |0>
do mu=1,n_det
! get the string of the determinant
call det_extract(det_mu,mu,N_int)
do ispin=1,2
! do the monoexcitation pq on it
call det_copy(det_mu,det_mu_pq,N_int)
call do_signed_mono_excitation(det_mu,det_mu_pq,mu_pq &
,ihole,ipart,ispin,phase,mu_pq_possible)
if (mu_pq_possible.eq.1) then
! possible, but not necessarily in the list
! do the second excitation
do jspin=1,2
call det_copy(det_mu_pq,det_mu_pqrs,N_int)
call do_signed_mono_excitation(det_mu_pq,det_mu_pqrs,mu_pqrs&
,jhole,jpart,jspin,phase2,mu_pqrs_possible)
! excitation possible
if (mu_pqrs_possible.eq.1) then
call i_H_psi(det_mu_pqrs,psi_det,psi_coef,N_int &
,N_det,N_det,N_states,i_H_psi_array)
do istate=1,N_states
res+=i_H_psi_array(istate)*psi_coef(mu,istate)*phase*phase2
end do
end if
! try the de-excitation with opposite sign
call det_copy(det_mu_pq,det_mu_pqrs,N_int)
call do_signed_mono_excitation(det_mu_pq,det_mu_pqrs,mu_pqrs&
,jpart,jhole,jspin,phase2,mu_pqrs_possible)
phase2=-phase2
! excitation possible
if (mu_pqrs_possible.eq.1) then
call i_H_psi(det_mu_pqrs,psi_det,psi_coef,N_int &
,N_det,N_det,N_states,i_H_psi_array)
do istate=1,N_states
res+=i_H_psi_array(istate)*psi_coef(mu,istate)*phase*phase2
end do
end if
end do
end if
! exchange the notion of pq and rs
! do the monoexcitation rs on the initial determinant
call det_copy(det_mu,det_mu_rs,N_int)
call do_signed_mono_excitation(det_mu,det_mu_rs,mu_rs &
,jhole,jpart,ispin,phase2,mu_rs_possible)
if (mu_rs_possible.eq.1) then
! do the second excitation
do jspin=1,2
call det_copy(det_mu_rs,det_mu_rspq,N_int)
call do_signed_mono_excitation(det_mu_rs,det_mu_rspq,mu_rspq&
,ihole,ipart,jspin,phase3,mu_rspq_possible)
! excitation possible (of course, the result is outside the CAS)
if (mu_rspq_possible.eq.1) then
call i_H_psi(det_mu_rspq,psi_det,psi_coef,N_int &
,N_det,N_det,N_states,i_H_psi_array)
do istate=1,N_states
res+=i_H_psi_array(istate)*psi_coef(mu,istate)*phase2*phase3
end do
end if
! we may try the de-excitation, with opposite sign
call det_copy(det_mu_rs,det_mu_rspq,N_int)
call do_signed_mono_excitation(det_mu_rs,det_mu_rspq,mu_rspq&
,ipart,ihole,jspin,phase3,mu_rspq_possible)
phase3=-phase3
! excitation possible (of course, the result is outside the CAS)
if (mu_rspq_possible.eq.1) then
call i_H_psi(det_mu_rspq,psi_det,psi_coef,N_int &
,N_det,N_det,N_states,i_H_psi_array)
do istate=1,N_states
res+=i_H_psi_array(istate)*psi_coef(mu,istate)*phase2*phase3
end do
end if
end do
end if
!
! the operator E H E, we have to do a double loop over the determinants
! we still have the determinant mu_pq and the phase in memory
if (mu_pq_possible.eq.1) then
do nu=1,N_det
call det_extract(det_nu,nu,N_int)
do jspin=1,2
call det_copy(det_nu,det_nu_rs,N_int)
call do_signed_mono_excitation(det_nu,det_nu_rs,nu_rs &
,jhole,jpart,jspin,phase2,nu_rs_possible)
! excitation possible ?
if (nu_rs_possible.eq.1) then
call i_H_j(det_mu_pq,det_nu_rs,N_int,i_H_j_element)
do istate=1,N_states
res+=2.D0*i_H_j_element*psi_coef(mu,istate) &
*psi_coef(nu,istate)*phase*phase2
end do
end if
end do
end do
end if
end do
end do
! state-averaged Hessian
res*=1.D0/dble(N_states)
end subroutine calc_hess_elem
BEGIN_PROVIDER [real*8, hessmat2, (nMonoEx,nMonoEx)]
BEGIN_DOC
! explicit hessian matrix from density matrices and integrals
! of course, this will be used for a direct Davidson procedure later
! we will not store the matrix in real life
! formulas are broken down as functions for the 6 classes of matrix elements
!
END_DOC
implicit none
integer :: i,j,t,u,a,b,indx,jndx,bstart,ustart,indx_shift
real*8 :: hessmat_itju
real*8 :: hessmat_itja
real*8 :: hessmat_itua
real*8 :: hessmat_iajb
real*8 :: hessmat_iatb
real*8 :: hessmat_taub
if (bavard) then
write(6,*) ' providing Hessian matrix hessmat2 '
write(6,*) ' nMonoEx = ',nMonoEx
endif
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP SHARED(hessmat2,n_core_inact_orb,n_act_orb,n_virt_orb,nMonoEx) &
!$OMP PRIVATE(i,indx,jndx,j,ustart,t,u,a,bstart,indx_shift)
!$OMP DO
do i=1,n_core_inact_orb
do t=1,n_act_orb
indx = t + (i-1)*n_act_orb
jndx=indx
do j=i,n_core_inact_orb
if (i.eq.j) then
ustart=t
else
ustart=1
end if
do u=ustart,n_act_orb
hessmat2(jndx,indx)=hessmat_itju(i,t,j,u)
jndx+=1
end do
end do
do j=1,n_core_inact_orb
do a=1,n_virt_orb
hessmat2(jndx,indx)=hessmat_itja(i,t,j,a)
jndx+=1
end do
end do
do u=1,n_act_orb
do a=1,n_virt_orb
hessmat2(jndx,indx)=hessmat_itua(i,t,u,a)
jndx+=1
end do
end do
end do
end do
!$OMP END DO NOWAIT
indx_shift = n_core_inact_orb*n_act_orb
!$OMP DO
do a=1,n_virt_orb
do i=1,n_core_inact_orb
indx = a + (i-1)*n_virt_orb + indx_shift
jndx=indx
do j=i,n_core_inact_orb
if (i.eq.j) then
bstart=a
else
bstart=1
end if
do b=bstart,n_virt_orb
hessmat2(jndx,indx)=hessmat_iajb(i,a,j,b)
jndx+=1
end do
end do
do t=1,n_act_orb
do b=1,n_virt_orb
hessmat2(jndx,indx)=hessmat_iatb(i,a,t,b)
jndx+=1
end do
end do
end do
end do
!$OMP END DO NOWAIT
indx_shift += n_core_inact_orb*n_virt_orb
!$OMP DO
do a=1,n_virt_orb
do t=1,n_act_orb
indx = a + (t-1)*n_virt_orb + indx_shift
jndx=indx
do u=t,n_act_orb
if (t.eq.u) then
bstart=a
else
bstart=1
end if
do b=bstart,n_virt_orb
hessmat2(jndx,indx)=hessmat_taub(t,a,u,b)
jndx+=1
end do
end do
end do
end do
!$OMP END DO
!$OMP END PARALLEL
do jndx=1,nMonoEx
do indx=1,jndx-1
hessmat2(indx,jndx) = hessmat2(jndx,indx)
enddo
enddo
END_PROVIDER
real*8 function hessmat_itju(i,t,j,u)
BEGIN_DOC
! the orbital hessian for core/inactive -> active, core/inactive -> active
! i, t, j, u are list indices, the corresponding orbitals are ii,tt,jj,uu
!
! we assume natural orbitals
END_DOC
implicit none
integer :: i,t,j,u,ii,tt,uu,v,vv,x,xx,y,jj
real*8 :: term,t2
ii=list_core_inact(i)
tt=list_act(t)
if (i.eq.j) then
if (t.eq.u) then
! diagonal element
term=occnum(tt)*Fipq(ii,ii)+2.D0*(Fipq(tt,tt)+Fapq(tt,tt)) &
-2.D0*(Fipq(ii,ii)+Fapq(ii,ii))
term+=2.D0*(3.D0*bielec_pxxq_no(tt,i,i,tt)-bielec_pqxx_no(tt,tt,i,i))
term-=2.D0*occnum(tt)*(3.D0*bielec_pxxq_no(tt,i,i,tt) &
-bielec_pqxx_no(tt,tt,i,i))
term-=occnum(tt)*Fipq(tt,tt)
do v=1,n_act_orb
vv=list_act(v)
do x=1,n_act_orb
xx=list_act(x)
term+=2.D0*(P0tuvx_no(t,t,v,x)*bielec_pqxx_no(vv,xx,i,i) &
+(P0tuvx_no(t,x,v,t)+P0tuvx_no(t,x,t,v))* &
bielec_pxxq_no(vv,i,i,xx))
do y=1,n_act_orb
term-=2.D0*P0tuvx_no(t,v,x,y)*bielecCI_no(t,v,y,xx)
end do
end do
end do
else
! it/iu, t != u
uu=list_act(u)
term=2.D0*(Fipq(tt,uu)+Fapq(tt,uu))
term+=2.D0*(4.D0*bielec_PxxQ_no(tt,i,j,uu)-bielec_PxxQ_no(uu,i,j,tt) &
-bielec_PQxx_no(tt,uu,i,j))
term-=occnum(tt)*Fipq(uu,tt)
term-=(occnum(tt)+occnum(uu)) &
*(3.D0*bielec_PxxQ_no(tt,i,i,uu)-bielec_PQxx_no(uu,tt,i,i))
do v=1,n_act_orb
vv=list_act(v)
! term-=D0tu(u,v)*Fipq(tt,vv) ! published, but inverting t and u seems more correct
do x=1,n_act_orb
xx=list_act(x)
term+=2.D0*(P0tuvx_no(u,t,v,x)*bielec_pqxx_no(vv,xx,i,i) &
+(P0tuvx_no(u,x,v,t)+P0tuvx_no(u,x,t,v)) &
*bielec_pxxq_no(vv,i,i,xx))
do y=1,n_act_orb
term-=2.D0*P0tuvx_no(t,v,x,y)*bielecCI_no(u,v,y,xx)
end do
end do
end do
end if
else
! it/ju
jj=list_core_inact(j)
uu=list_act(u)
if (t.eq.u) then
term=occnum(tt)*Fipq(ii,jj)
term-=2.D0*(Fipq(ii,jj)+Fapq(ii,jj))
else
term=0.D0
end if
term+=2.D0*(4.D0*bielec_PxxQ_no(tt,i,j,uu)-bielec_PxxQ_no(uu,i,j,tt) &
-bielec_PQxx_no(tt,uu,i,j))
term-=(occnum(tt)+occnum(uu))* &
(4.D0*bielec_PxxQ_no(tt,i,j,uu)-bielec_PxxQ_no(uu,i,j,tt) &
-bielec_PQxx_no(uu,tt,i,j))
do v=1,n_act_orb
vv=list_act(v)
do x=1,n_act_orb
xx=list_act(x)
term+=2.D0*(P0tuvx_no(u,t,v,x)*bielec_pqxx_no(vv,xx,i,j) &
+(P0tuvx_no(u,x,v,t)+P0tuvx_no(u,x,t,v)) &
*bielec_pxxq_no(vv,i,j,xx))
end do
end do
end if
term*=2.D0
hessmat_itju=term
end function hessmat_itju
real*8 function hessmat_itja(i,t,j,a)
BEGIN_DOC
! the orbital hessian for core/inactive -> active, core/inactive -> virtual
END_DOC
implicit none
integer :: i,t,j,a,ii,tt,jj,aa,v,vv,x,y
real*8 :: term
! it/ja
ii=list_core_inact(i)
tt=list_act(t)
jj=list_core_inact(j)
aa=list_virt(a)
term=2.D0*(4.D0*bielec_pxxq_no(aa,j,i,tt) &
-bielec_pqxx_no(aa,tt,i,j) -bielec_pxxq_no(aa,i,j,tt))
term-=occnum(tt)*(4.D0*bielec_pxxq_no(aa,j,i,tt) &
-bielec_pqxx_no(aa,tt,i,j) -bielec_pxxq_no(aa,i,j,tt))
if (i.eq.j) then
term+=2.D0*(Fipq(aa,tt)+Fapq(aa,tt))
term-=0.5D0*occnum(tt)*Fipq(aa,tt)
do v=1,n_act_orb
do x=1,n_act_orb
do y=1,n_act_orb
term-=P0tuvx_no(t,v,x,y)*bielecCI_no(x,y,v,aa)
end do
end do
end do
end if
term*=2.D0
hessmat_itja=term
end function hessmat_itja
real*8 function hessmat_itua(i,t,u,a)
BEGIN_DOC
! the orbital hessian for core/inactive -> active, active -> virtual
END_DOC
implicit none
integer :: i,t,u,a,ii,tt,uu,aa,v,vv,x,xx,u3,t3,v3
real*8 :: term
ii=list_core_inact(i)
tt=list_act(t)
t3=t+n_core_inact_orb
uu=list_act(u)
u3=u+n_core_inact_orb
aa=list_virt(a)
if (t.eq.u) then
term=-occnum(tt)*Fipq(aa,ii)
else
term=0.D0
end if
term-=occnum(uu)*(bielec_pqxx_no(aa,ii,t3,u3)-4.D0*bielec_pqxx_no(aa,uu,t3,i)&
+bielec_pxxq_no(aa,t3,u3,ii))
do v=1,n_act_orb
vv=list_act(v)
v3=v+n_core_inact_orb
do x=1,n_act_orb
integer :: x3
xx=list_act(x)
x3=x+n_core_inact_orb
term-=2.D0*(P0tuvx_no(t,u,v,x)*bielec_pqxx_no(aa,ii,v3,x3) &
+(P0tuvx_no(t,v,u,x)+P0tuvx_no(t,v,x,u)) &
*bielec_pqxx_no(aa,xx,v3,i))
end do
end do
if (t.eq.u) then
term+=Fipq(aa,ii)+Fapq(aa,ii)
end if
term*=2.D0
hessmat_itua=term
end function hessmat_itua
real*8 function hessmat_iajb(i,a,j,b)
BEGIN_DOC
! the orbital hessian for core/inactive -> virtual, core/inactive -> virtual
END_DOC
implicit none
integer :: i,a,j,b,ii,aa,jj,bb
real*8 :: term
ii=list_core_inact(i)
aa=list_virt(a)
if (i.eq.j) then
if (a.eq.b) then
! ia/ia
term=2.D0*(Fipq(aa,aa)+Fapq(aa,aa)-Fipq(ii,ii)-Fapq(ii,ii))
term+=2.D0*(3.D0*bielec_pxxq_no(aa,i,i,aa)-bielec_pqxx_no(aa,aa,i,i))
else
bb=list_virt(b)
! ia/ib
term=2.D0*(Fipq(aa,bb)+Fapq(aa,bb))
term+=2.D0*(3.D0*bielec_pxxq_no(aa,i,i,bb)-bielec_pqxx_no(aa,bb,i,i))
end if
else
! ia/jb
jj=list_core_inact(j)
bb=list_virt(b)
term=2.D0*(4.D0*bielec_pxxq_no(aa,i,j,bb)-bielec_pqxx_no(aa,bb,i,j) &
-bielec_pxxq_no(aa,j,i,bb))
if (a.eq.b) then
term-=2.D0*(Fipq(ii,jj)+Fapq(ii,jj))
end if
end if
term*=2.D0
hessmat_iajb=term
end function hessmat_iajb
real*8 function hessmat_iatb(i,a,t,b)
BEGIN_DOC
! the orbital hessian for core/inactive -> virtual, active -> virtual
END_DOC
implicit none
integer :: i,a,t,b,ii,aa,tt,bb,v,vv,x,y,v3,t3
real*8 :: term
ii=list_core_inact(i)
aa=list_virt(a)
tt=list_act(t)
bb=list_virt(b)
t3=t+n_core_inact_orb
term=occnum(tt)*(4.D0*bielec_pxxq_no(aa,i,t3,bb)-bielec_pxxq_no(aa,t3,i,bb)&
-bielec_pqxx_no(aa,bb,i,t3))
if (a.eq.b) then
term-=Fipq(tt,ii)+Fapq(tt,ii)
term-=0.5D0*occnum(tt)*Fipq(tt,ii)
do v=1,n_act_orb
do x=1,n_act_orb
do y=1,n_act_orb
term-=P0tuvx_no(t,v,x,y)*bielecCI_no(x,y,v,ii)
end do
end do
end do
end if
term*=2.D0
hessmat_iatb=term
end function hessmat_iatb
real*8 function hessmat_taub(t,a,u,b)
BEGIN_DOC
! the orbital hessian for act->virt,act->virt
END_DOC
implicit none
integer :: t,a,u,b,tt,aa,uu,bb,v,vv,x,xx,y
integer :: v3,x3
real*8 :: term,t1,t2,t3
tt=list_act(t)
aa=list_virt(a)
if (t == u) then
if (a == b) then
! ta/ta
t1=occnum(tt)*Fipq(aa,aa)
t2=0.D0
t3=0.D0
t1-=occnum(tt)*Fipq(tt,tt)
do v=1,n_act_orb
vv=list_act(v)
v3=v+n_core_inact_orb
do x=1,n_act_orb
xx=list_act(x)
x3=x+n_core_inact_orb
t2+=2.D0*(P0tuvx_no(t,t,v,x)*bielec_pqxx_no(aa,aa,v3,x3) &
+(P0tuvx_no(t,x,v,t)+P0tuvx_no(t,x,t,v))* &
bielec_pxxq_no(aa,x3,v3,aa))
do y=1,n_act_orb
t3-=2.D0*P0tuvx_no(t,v,x,y)*bielecCI_no(t,v,y,xx)
end do
end do
end do
term=t1+t2+t3
else
bb=list_virt(b)
! ta/tb b/=a
term=occnum(tt)*Fipq(aa,bb)
do v=1,n_act_orb
vv=list_act(v)
v3=v+n_core_inact_orb
do x=1,n_act_orb
xx=list_act(x)
x3=x+n_core_inact_orb
term+=2.D0*(P0tuvx_no(t,t,v,x)*bielec_pqxx_no(aa,bb,v3,x3) &
+(P0tuvx_no(t,x,v,t)+P0tuvx_no(t,x,t,v)) &
*bielec_pxxq_no(aa,x3,v3,bb))
end do
end do
end if
else
! ta/ub t/=u
uu=list_act(u)
bb=list_virt(b)
term=0.D0
do v=1,n_act_orb
vv=list_act(v)
v3=v+n_core_inact_orb
do x=1,n_act_orb
xx=list_act(x)
x3=x+n_core_inact_orb
term+=2.D0*(P0tuvx_no(t,u,v,x)*bielec_pqxx_no(aa,bb,v3,x3) &
+(P0tuvx_no(t,x,v,u)+P0tuvx_no(t,x,u,v)) &
*bielec_pxxq_no(aa,x3,v3,bb))
end do
end do
if (a.eq.b) then
term-=0.5D0*(occnum(tt)*Fipq(uu,tt)+occnum(uu)*Fipq(tt,uu))
do v=1,n_act_orb
do y=1,n_act_orb
do x=1,n_act_orb
term-=P0tuvx_no(t,v,x,y)*bielecCI_no(x,y,v,uu)
term-=P0tuvx_no(u,v,x,y)*bielecCI_no(x,y,v,tt)
end do
end do
end do
end if
end if
term*=2.D0
hessmat_taub=term
end function hessmat_taub
BEGIN_PROVIDER [real*8, hessdiag, (nMonoEx)]
BEGIN_DOC
! the diagonal of the Hessian, needed for the Davidson procedure
END_DOC
implicit none
integer :: i,t,a,indx,indx_shift
real*8 :: hessmat_itju,hessmat_iajb,hessmat_taub
!$OMP PARALLEL DEFAULT(NONE) &
!$OMP SHARED(hessdiag,n_core_inact_orb,n_act_orb,n_virt_orb,nMonoEx) &
!$OMP PRIVATE(i,indx,t,a,indx_shift)
!$OMP DO
do i=1,n_core_inact_orb
do t=1,n_act_orb
indx = t + (i-1)*n_act_orb
hessdiag(indx)=hessmat_itju(i,t,i,t)
end do
end do
!$OMP END DO NOWAIT
indx_shift = n_core_inact_orb*n_act_orb
!$OMP DO
do a=1,n_virt_orb
do i=1,n_core_inact_orb
indx = a + (i-1)*n_virt_orb + indx_shift
hessdiag(indx)=hessmat_iajb(i,a,i,a)
end do
end do
!$OMP END DO NOWAIT
indx_shift += n_core_inact_orb*n_virt_orb
!$OMP DO
do a=1,n_virt_orb
do t=1,n_act_orb
indx = a + (t-1)*n_virt_orb + indx_shift
hessdiag(indx)=hessmat_taub(t,a,t,a)
end do
end do
!$OMP END DO
!$OMP END PARALLEL
END_PROVIDER

View File

@ -1,80 +0,0 @@
BEGIN_PROVIDER [real*8, Fipq, (mo_num,mo_num) ]
BEGIN_DOC
! the inactive Fock matrix, in molecular orbitals
END_DOC
implicit none
integer :: p,q,k,kk,t,tt,u,uu
do q=1,mo_num
do p=1,mo_num
Fipq(p,q)=one_ints_no(p,q)
end do
end do
! the inactive Fock matrix
do k=1,n_core_inact_orb
kk=list_core_inact(k)
do q=1,mo_num
do p=1,mo_num
Fipq(p,q)+=2.D0*bielec_pqxx_no(p,q,k,k) -bielec_pxxq_no(p,k,k,q)
end do
end do
end do
if (bavard) then
integer :: i
write(6,*)
write(6,*) ' the diagonal of the inactive effective Fock matrix '
write(6,'(5(i3,F12.5))') (i,Fipq(i,i),i=1,mo_num)
write(6,*)
end if
END_PROVIDER
BEGIN_PROVIDER [real*8, Fapq, (mo_num,mo_num) ]
BEGIN_DOC
! the active active Fock matrix, in molecular orbitals
! we create them in MOs, quite expensive
!
! for an implementation in AOs we need first the natural orbitals
! for forming an active density matrix in AOs
!
END_DOC
implicit none
integer :: p,q,k,kk,t,tt,u,uu
Fapq = 0.d0
! the active Fock matrix, D0tu is diagonal
do t=1,n_act_orb
tt=list_act(t)
do q=1,mo_num
do p=1,mo_num
Fapq(p,q)+=occnum(tt) &
*(bielec_pqxx_no(p,q,tt,tt)-0.5D0*bielec_pxxq_no(p,tt,tt,q))
end do
end do
end do
if (bavard) then
integer :: i
write(6,*)
write(6,*) ' the effective Fock matrix over MOs'
write(6,*)
write(6,*)
write(6,*) ' the diagonal of the inactive effective Fock matrix '
write(6,'(5(i3,F12.5))') (i,Fipq(i,i),i=1,mo_num)
write(6,*)
write(6,*)
write(6,*) ' the diagonal of the active Fock matrix '
write(6,'(5(i3,F12.5))') (i,Fapq(i,i),i=1,mo_num)
write(6,*)
end if
END_PROVIDER

View File

@ -1,231 +0,0 @@
BEGIN_PROVIDER [real*8, occnum, (mo_num)]
implicit none
BEGIN_DOC
! MO occupation numbers
END_DOC
integer :: i
occnum=0.D0
do i=1,n_core_inact_orb
occnum(list_core_inact(i))=2.D0
end do
do i=1,n_act_orb
occnum(list_act(i))=occ_act(i)
end do
if (bavard) then
write(6,*) ' occupation numbers '
do i=1,mo_num
write(6,*) i,occnum(i)
end do
endif
END_PROVIDER
BEGIN_PROVIDER [ real*8, natorbsCI, (n_act_orb,n_act_orb) ]
&BEGIN_PROVIDER [ real*8, occ_act, (n_act_orb) ]
implicit none
BEGIN_DOC
! Natural orbitals of CI
END_DOC
integer :: i, j
double precision :: Vt(n_act_orb,n_act_orb)
! call lapack_diag(occ_act,natorbsCI,D0tu,n_act_orb,n_act_orb)
call svd(D0tu, size(D0tu,1), natorbsCI,size(natorbsCI,1), occ_act, Vt, size(Vt,1),n_act_orb,n_act_orb)
if (bavard) then
write(6,*) ' found occupation numbers as '
do i=1,n_act_orb
write(6,*) i,occ_act(i)
end do
integer :: nmx
real*8 :: xmx
do i=1,n_act_orb
! largest element of the eigenvector should be positive
xmx=0.D0
nmx=0
do j=1,n_act_orb
if (abs(natOrbsCI(j,i)).gt.xmx) then
nmx=j
xmx=abs(natOrbsCI(j,i))
end if
end do
xmx=sign(1.D0,natOrbsCI(nmx,i))
do j=1,n_act_orb
natOrbsCI(j,i)*=xmx
end do
write(6,*) ' Eigenvector No ',i
write(6,'(5(I3,F12.5))') (j,natOrbsCI(j,i),j=1,n_act_orb)
end do
end if
END_PROVIDER
BEGIN_PROVIDER [real*8, P0tuvx_no, (n_act_orb,n_act_orb,n_act_orb,n_act_orb)]
implicit none
BEGIN_DOC
! 4-index transformation of 2part matrices
END_DOC
integer :: i,j,k,l,p,q
real*8 :: d(n_act_orb)
! index per index
! first quarter
P0tuvx_no(:,:,:,:) = P0tuvx(:,:,:,:)
do j=1,n_act_orb
do k=1,n_act_orb
do l=1,n_act_orb
do p=1,n_act_orb
d(p)=0.D0
end do
do p=1,n_act_orb
do q=1,n_act_orb
d(p)+=P0tuvx_no(q,j,k,l)*natorbsCI(q,p)
end do
end do
do p=1,n_act_orb
P0tuvx_no(p,j,k,l)=d(p)
end do
end do
end do
end do
! 2nd quarter
do j=1,n_act_orb
do k=1,n_act_orb
do l=1,n_act_orb
do p=1,n_act_orb
d(p)=0.D0
end do
do p=1,n_act_orb
do q=1,n_act_orb
d(p)+=P0tuvx_no(j,q,k,l)*natorbsCI(q,p)
end do
end do
do p=1,n_act_orb
P0tuvx_no(j,p,k,l)=d(p)
end do
end do
end do
end do
! 3rd quarter
do j=1,n_act_orb
do k=1,n_act_orb
do l=1,n_act_orb
do p=1,n_act_orb
d(p)=0.D0
end do
do p=1,n_act_orb
do q=1,n_act_orb
d(p)+=P0tuvx_no(j,k,q,l)*natorbsCI(q,p)
end do
end do
do p=1,n_act_orb
P0tuvx_no(j,k,p,l)=d(p)
end do
end do
end do
end do
! 4th quarter
do j=1,n_act_orb
do k=1,n_act_orb
do l=1,n_act_orb
do p=1,n_act_orb
d(p)=0.D0
end do
do p=1,n_act_orb
do q=1,n_act_orb
d(p)+=P0tuvx_no(j,k,l,q)*natorbsCI(q,p)
end do
end do
do p=1,n_act_orb
P0tuvx_no(j,k,l,p)=d(p)
end do
end do
end do
end do
END_PROVIDER
BEGIN_PROVIDER [real*8, one_ints_no, (mo_num,mo_num)]
implicit none
BEGIN_DOC
! Transformed one-e integrals
END_DOC
integer :: i,j, p, q
real*8 :: d(n_act_orb)
one_ints_no(:,:)=mo_one_e_integrals(:,:)
! 1st half-trf
do j=1,mo_num
do p=1,n_act_orb
d(p)=0.D0
end do
do p=1,n_act_orb
do q=1,n_act_orb
d(p)+=one_ints_no(list_act(q),j)*natorbsCI(q,p)
end do
end do
do p=1,n_act_orb
one_ints_no(list_act(p),j)=d(p)
end do
end do
! 2nd half-trf
do j=1,mo_num
do p=1,n_act_orb
d(p)=0.D0
end do
do p=1,n_act_orb
do q=1,n_act_orb
d(p)+=one_ints_no(j,list_act(q))*natorbsCI(q,p)
end do
end do
do p=1,n_act_orb
one_ints_no(j,list_act(p))=d(p)
end do
end do
END_PROVIDER
BEGIN_PROVIDER [ double precision, NatOrbsCI_mos, (mo_num, mo_num) ]
implicit none
BEGIN_DOC
! Rotation matrix from current MOs to the CI natural MOs
END_DOC
integer :: p,q
NatOrbsCI_mos(:,:) = 0.d0
do q = 1,mo_num
NatOrbsCI_mos(q,q) = 1.d0
enddo
do q = 1,n_act_orb
do p = 1,n_act_orb
NatOrbsCI_mos(list_act(p),list_act(q)) = natorbsCI(p,q)
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [real*8, NatOrbsFCI, (ao_num,mo_num)]
implicit none
BEGIN_DOC
! FCI natural orbitals
END_DOC
call dgemm('N','N', ao_num,mo_num,mo_num,1.d0, &
mo_coef, size(mo_coef,1), &
NatOrbsCI_mos, size(NatOrbsCI_mos,1), 0.d0, &
NatOrbsFCI, size(NatOrbsFCI,1))
END_PROVIDER

View File

@ -1,221 +0,0 @@
BEGIN_PROVIDER [real*8, SXmatrix, (nMonoEx+1,nMonoEx+1)]
implicit none
BEGIN_DOC
! Single-excitation matrix
END_DOC
integer :: i,j
do i=1,nMonoEx+1
do j=1,nMonoEx+1
SXmatrix(i,j)=0.D0
end do
end do
do i=1,nMonoEx
SXmatrix(1,i+1)=gradvec2(i)
SXmatrix(1+i,1)=gradvec2(i)
end do
do i=1,nMonoEx
do j=1,nMonoEx
SXmatrix(i+1,j+1)=hessmat2(i,j)
SXmatrix(j+1,i+1)=hessmat2(i,j)
end do
end do
do i = 1, nMonoEx
SXmatrix(i+1,i+1) += level_shift_casscf
enddo
if (bavard) then
do i=2,nMonoEx
write(6,*) ' diagonal of the Hessian : ',i,hessmat2(i,i)
end do
end if
END_PROVIDER
BEGIN_PROVIDER [real*8, SXeigenvec, (nMonoEx+1,nMonoEx+1)]
&BEGIN_PROVIDER [real*8, SXeigenval, (nMonoEx+1)]
implicit none
BEGIN_DOC
! Eigenvectors/eigenvalues of the single-excitation matrix
END_DOC
call lapack_diag(SXeigenval,SXeigenvec,SXmatrix,nMonoEx+1,nMonoEx+1)
if (bavard) then
write(6,*) ' SXdiag : lowest 5 eigenvalues '
write(6,*) ' 1 - ',SXeigenval(1),SXeigenvec(1,1)
if(nmonoex.gt.0)then
write(6,*) ' 2 - ',SXeigenval(2),SXeigenvec(1,2)
write(6,*) ' 3 - ',SXeigenval(3),SXeigenvec(1,3)
write(6,*) ' 4 - ',SXeigenval(4),SXeigenvec(1,4)
write(6,*) ' 5 - ',SXeigenval(5),SXeigenvec(1,5)
endif
write(6,*)
write(6,*) ' SXdiag : lowest eigenvalue = ',SXeigenval(1)
endif
END_PROVIDER
BEGIN_PROVIDER [real*8, energy_improvement]
implicit none
if(state_following_casscf)then
energy_improvement = SXeigenval(best_vector_ovrlp_casscf)
else
energy_improvement = SXeigenval(1)
endif
END_PROVIDER
BEGIN_PROVIDER [ integer, best_vector_ovrlp_casscf ]
&BEGIN_PROVIDER [ double precision, best_overlap_casscf ]
implicit none
integer :: i
double precision :: c0
best_overlap_casscf = 0.D0
best_vector_ovrlp_casscf = -1000
do i=1,nMonoEx+1
if (SXeigenval(i).lt.0.D0) then
if (abs(SXeigenvec(1,i)).gt.best_overlap_casscf) then
best_overlap_casscf=abs(SXeigenvec(1,i))
best_vector_ovrlp_casscf = i
end if
end if
end do
if(best_vector_ovrlp_casscf.lt.0)then
best_vector_ovrlp_casscf = minloc(SXeigenval,nMonoEx+1)
endif
c0=SXeigenvec(1,best_vector_ovrlp_casscf)
if (bavard) then
write(6,*) ' SXdiag : eigenvalue for best overlap with '
write(6,*) ' previous orbitals = ',SXeigenval(best_vector_ovrlp_casscf)
write(6,*) ' weight of the 1st element ',c0
endif
END_PROVIDER
BEGIN_PROVIDER [double precision, SXvector, (nMonoEx+1)]
implicit none
BEGIN_DOC
! Best eigenvector of the single-excitation matrix
END_DOC
integer :: i
double precision :: c0
c0=SXeigenvec(1,best_vector_ovrlp_casscf)
do i=1,nMonoEx+1
SXvector(i)=SXeigenvec(i,best_vector_ovrlp_casscf)/c0
end do
END_PROVIDER
BEGIN_PROVIDER [double precision, NewOrbs, (ao_num,mo_num) ]
implicit none
BEGIN_DOC
! Updated orbitals
END_DOC
integer :: i,j,ialph
if(state_following_casscf)then
print*,'Using the state following casscf '
call dgemm('N','T', ao_num,mo_num,mo_num,1.d0, &
NatOrbsFCI, size(NatOrbsFCI,1), &
Umat, size(Umat,1), 0.d0, &
NewOrbs, size(NewOrbs,1))
level_shift_casscf *= 0.5D0
level_shift_casscf = max(level_shift_casscf,0.002d0)
!touch level_shift_casscf
else
if(best_vector_ovrlp_casscf.ne.1.and.n_orb_swap.ne.0)then
print*,'Taking the lowest root for the CASSCF'
print*,'!!! SWAPPING MOS !!!!!!'
level_shift_casscf *= 2.D0
level_shift_casscf = min(level_shift_casscf,0.5d0)
print*,'level_shift_casscf = ',level_shift_casscf
NewOrbs = switch_mo_coef
!mo_coef = switch_mo_coef
!soft_touch mo_coef
!call save_mos_no_occ
!stop
else
level_shift_casscf *= 0.5D0
level_shift_casscf = max(level_shift_casscf,0.002d0)
!touch level_shift_casscf
call dgemm('N','T', ao_num,mo_num,mo_num,1.d0, &
NatOrbsFCI, size(NatOrbsFCI,1), &
Umat, size(Umat,1), 0.d0, &
NewOrbs, size(NewOrbs,1))
endif
endif
END_PROVIDER
BEGIN_PROVIDER [real*8, Umat, (mo_num,mo_num) ]
implicit none
BEGIN_DOC
! Orbital rotation matrix
END_DOC
integer :: i,j,indx,k,iter,t,a,ii,tt,aa
logical :: converged
real*8 :: Tpotmat (mo_num,mo_num), Tpotmat2 (mo_num,mo_num)
real*8 :: Tmat(mo_num,mo_num)
real*8 :: f
! the orbital rotation matrix T
Tmat(:,:)=0.D0
indx=1
do i=1,n_core_inact_orb
ii=list_core_inact(i)
do t=1,n_act_orb
tt=list_act(t)
indx+=1
Tmat(ii,tt)= SXvector(indx)
Tmat(tt,ii)=-SXvector(indx)
end do
end do
do i=1,n_core_inact_orb
ii=list_core_inact(i)
do a=1,n_virt_orb
aa=list_virt(a)
indx+=1
Tmat(ii,aa)= SXvector(indx)
Tmat(aa,ii)=-SXvector(indx)
end do
end do
do t=1,n_act_orb
tt=list_act(t)
do a=1,n_virt_orb
aa=list_virt(a)
indx+=1
Tmat(tt,aa)= SXvector(indx)
Tmat(aa,tt)=-SXvector(indx)
end do
end do
! Form the exponential
Tpotmat(:,:)=0.D0
Umat(:,:) =0.D0
do i=1,mo_num
Tpotmat(i,i)=1.D0
Umat(i,i) =1.d0
end do
iter=0
converged=.false.
do while (.not.converged)
iter+=1
f = 1.d0 / dble(iter)
Tpotmat2(:,:) = Tpotmat(:,:) * f
call dgemm('N','N', mo_num,mo_num,mo_num,1.d0, &
Tpotmat2, size(Tpotmat2,1), &
Tmat, size(Tmat,1), 0.d0, &
Tpotmat, size(Tpotmat,1))
Umat(:,:) = Umat(:,:) + Tpotmat(:,:)
converged = ( sum(abs(Tpotmat(:,:))) < 1.d-6).or.(iter>30)
end do
END_PROVIDER

View File

@ -1,70 +0,0 @@
subroutine reorder_orbitals_for_casscf
implicit none
BEGIN_DOC
! routine that reorders the orbitals of the CASSCF in terms block of core, active and virtual
END_DOC
integer :: i,j,iorb
integer, allocatable :: iorder(:),array(:)
allocate(iorder(mo_num),array(mo_num))
do i = 1, n_core_orb
iorb = list_core(i)
array(iorb) = i
enddo
do i = 1, n_inact_orb
iorb = list_inact(i)
array(iorb) = mo_num + i
enddo
do i = 1, n_act_orb
iorb = list_act(i)
array(iorb) = 2 * mo_num + i
enddo
do i = 1, n_virt_orb
iorb = list_virt(i)
array(iorb) = 3 * mo_num + i
enddo
do i = 1, mo_num
iorder(i) = i
enddo
call isort(array,iorder,mo_num)
double precision, allocatable :: mo_coef_new(:,:)
allocate(mo_coef_new(ao_num,mo_num))
do i = 1, mo_num
mo_coef_new(:,i) = mo_coef(:,iorder(i))
enddo
mo_coef = mo_coef_new
touch mo_coef
list_core_reverse = 0
do i = 1, n_core_orb
list_core(i) = i
list_core_reverse(i) = i
mo_class(i) = "Core"
enddo
list_inact_reverse = 0
do i = 1, n_inact_orb
list_inact(i) = i + n_core_orb
list_inact_reverse(i+n_core_orb) = i
mo_class(i+n_core_orb) = "Inactive"
enddo
list_act_reverse = 0
do i = 1, n_act_orb
list_act(i) = n_core_inact_orb + i
list_act_reverse(n_core_inact_orb + i) = i
mo_class(n_core_inact_orb + i) = "Active"
enddo
list_virt_reverse = 0
do i = 1, n_virt_orb
list_virt(i) = n_core_inact_orb + n_act_orb + i
list_virt_reverse(n_core_inact_orb + n_act_orb + i) = i
mo_class(n_core_inact_orb + n_act_orb + i) = "Virtual"
enddo
touch list_core_reverse list_core list_inact list_inact_reverse list_act list_act_reverse list_virt list_virt_reverse
end

View File

@ -1,9 +0,0 @@
subroutine save_energy(E,pt2)
implicit none
BEGIN_DOC
! Saves the energy in |EZFIO|.
END_DOC
double precision, intent(in) :: E(N_states), pt2(N_states)
call ezfio_set_casscf_energy(E(1:N_states))
call ezfio_set_casscf_energy_pt2(E(1:N_states)+pt2(1:N_states))
end

View File

@ -1,207 +0,0 @@
BEGIN_PROVIDER [double precision, super_ci_dm, (mo_num,mo_num)]
implicit none
BEGIN_DOC
! density matrix of the super CI matrix, in the basis of NATURAL ORBITALS OF THE CASCI WF
!
! This is obtained from annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
!
! WARNING ::: in the equation B3.d there is a TYPO with a forgotten MINUS SIGN (see variable mat_tmp_dm_super_ci )
END_DOC
super_ci_dm = 0.d0
integer :: i,j,iorb,jorb
integer :: a,aorb,b,borb
integer :: t,torb,v,vorb,u,uorb,x,xorb
double precision :: c0,ci
c0 = SXeigenvec(1,1)
! equation B3.a of the annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
! loop over the core/inact
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
super_ci_dm(iorb,iorb) = 2.d0 ! first term of B3.a
! loop over the core/inact
do j = 1, n_core_inact_orb
jorb = list_core_inact(j)
! loop over the virtual
do a = 1, n_virt_orb
aorb = list_virt(a)
super_ci_dm(jorb,iorb) += -2.d0 * lowest_super_ci_coef_mo(aorb,iorb) * lowest_super_ci_coef_mo(aorb,jorb) ! second term in B3.a
enddo
do t = 1, n_act_orb
torb = list_act(t)
! thrid term of the B3.a
super_ci_dm(jorb,iorb) += - lowest_super_ci_coef_mo(iorb,torb) * lowest_super_ci_coef_mo(jorb,torb) * (2.d0 - occ_act(t))
enddo
enddo
enddo
! equation B3.b of the annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
do t = 1, n_act_orb
torb = list_act(t)
super_ci_dm(iorb,torb) = c0 * lowest_super_ci_coef_mo(torb,iorb) * (2.d0 - occ_act(t))
super_ci_dm(torb,iorb) = c0 * lowest_super_ci_coef_mo(torb,iorb) * (2.d0 - occ_act(t))
do a = 1, n_virt_orb
aorb = list_virt(a)
super_ci_dm(iorb,torb) += - lowest_super_ci_coef_mo(aorb,iorb) * lowest_super_ci_coef_mo(aorb,torb) * occ_act(t)
super_ci_dm(torb,iorb) += - lowest_super_ci_coef_mo(aorb,iorb) * lowest_super_ci_coef_mo(aorb,torb) * occ_act(t)
enddo
enddo
enddo
! equation B3.c of the annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
do a = 1, n_virt_orb
aorb = list_virt(a)
super_ci_dm(aorb,iorb) = 2.d0 * c0 * lowest_super_ci_coef_mo(aorb,iorb)
super_ci_dm(iorb,aorb) = 2.d0 * c0 * lowest_super_ci_coef_mo(aorb,iorb)
enddo
enddo
! equation B3.d of the annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
do t = 1, n_act_orb
torb = list_act(t)
super_ci_dm(torb,torb) = occ_act(t) ! first term of equation B3.d
do x = 1, n_act_orb
xorb = list_act(x)
super_ci_dm(torb,torb) += - occ_act(x) * occ_act(t)* mat_tmp_dm_super_ci(x,x) ! second term involving the ONE-rdm
enddo
do u = 1, n_act_orb
uorb = list_act(u)
! second term of equation B3.d
do x = 1, n_act_orb
xorb = list_act(x)
do v = 1, n_act_orb
vorb = list_act(v)
super_ci_dm(torb,uorb) += 2.d0 * P0tuvx_no(v,x,t,u) * mat_tmp_dm_super_ci(v,x) ! second term involving the TWO-rdm
enddo
enddo
! third term of equation B3.d
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
super_ci_dm(torb,uorb) += lowest_super_ci_coef_mo(iorb,torb) * lowest_super_ci_coef_mo(iorb,uorb) * (2.d0 - occ_act(t) - occ_act(u))
enddo
enddo
enddo
! equation B3.e of the annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
do t = 1, n_act_orb
torb = list_act(t)
do a = 1, n_virt_orb
aorb = list_virt(a)
super_ci_dm(aorb,torb) += c0 * lowest_super_ci_coef_mo(aorb,torb) * occ_act(t)
super_ci_dm(torb,aorb) += c0 * lowest_super_ci_coef_mo(aorb,torb) * occ_act(t)
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
super_ci_dm(aorb,torb) += lowest_super_ci_coef_mo(iorb,aorb) * lowest_super_ci_coef_mo(iorb,torb) * (2.d0 - occ_act(t))
super_ci_dm(torb,aorb) += lowest_super_ci_coef_mo(iorb,aorb) * lowest_super_ci_coef_mo(iorb,torb) * (2.d0 - occ_act(t))
enddo
enddo
enddo
! equation B3.f of the annex B of Roos et. al. Chemical Physics 48 (1980) 157-173
do a = 1, n_virt_orb
aorb = list_virt(a)
do b = 1, n_virt_orb
borb= list_virt(b)
! First term of equation B3.f
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
super_ci_dm(borb,aorb) += 2.d0 * lowest_super_ci_coef_mo(iorb,aorb) * lowest_super_ci_coef_mo(iorb,borb)
enddo
! Second term of equation B3.f
do t = 1, n_act_orb
torb = list_act(t)
super_ci_dm(borb,aorb) += lowest_super_ci_coef_mo(torb,aorb) * lowest_super_ci_coef_mo(torb,borb) * occ_act(t)
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [double precision, superci_natorb, (ao_num,mo_num)
&BEGIN_PROVIDER [double precision, superci_nat_occ, (mo_num)
implicit none
call general_mo_coef_new_as_svd_vectors_of_mo_matrix_eig(super_ci_dm,mo_num,mo_num,mo_num,NatOrbsFCI,superci_nat_occ,superci_natorb)
END_PROVIDER
BEGIN_PROVIDER [double precision, mat_tmp_dm_super_ci, (n_act_orb,n_act_orb)]
implicit none
BEGIN_DOC
! computation of the term in [ ] in the equation B3.d of Roos et. al. Chemical Physics 48 (1980) 157-173
!
! !!!!! WARNING !!!!!! there is a TYPO: a MINUS SIGN SHOULD APPEAR in that term
END_DOC
integer :: a,aorb,i,iorb
integer :: x,xorb,v,vorb
mat_tmp_dm_super_ci = 0.d0
do v = 1, n_act_orb
vorb = list_act(v)
do x = 1, n_act_orb
xorb = list_act(x)
do a = 1, n_virt_orb
aorb = list_virt(a)
mat_tmp_dm_super_ci(x,v) += lowest_super_ci_coef_mo(aorb,vorb) * lowest_super_ci_coef_mo(aorb,xorb)
enddo
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
! MARK THE MINUS SIGN HERE !!!!!!!!!!! BECAUSE OF TYPO IN THE ORIGINAL PAPER
mat_tmp_dm_super_ci(x,v) -= lowest_super_ci_coef_mo(iorb,vorb) * lowest_super_ci_coef_mo(iorb,xorb)
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [double precision, lowest_super_ci_coef_mo, (mo_num,mo_num)]
implicit none
integer :: i,j,iorb,jorb
integer :: a, aorb,t, torb
double precision :: sqrt2
sqrt2 = 1.d0/dsqrt(2.d0)
do i = 1, nMonoEx
iorb = excit(1,i)
jorb = excit(2,i)
lowest_super_ci_coef_mo(iorb,jorb) = SXeigenvec(i+1,1)
lowest_super_ci_coef_mo(jorb,iorb) = SXeigenvec(i+1,1)
enddo
! a_{it} of the equation B.2 of Roos et. al. Chemical Physics 48 (1980) 157-173
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
do t = 1, n_act_orb
torb = list_act(t)
lowest_super_ci_coef_mo(torb,iorb) *= (2.d0 - occ_act(t))**(-0.5d0)
lowest_super_ci_coef_mo(iorb,torb) *= (2.d0 - occ_act(t))**(-0.5d0)
enddo
enddo
! a_{ia} of the equation B.2 of Roos et. al. Chemical Physics 48 (1980) 157-173
do i = 1, n_core_inact_orb
iorb = list_core_inact(i)
do a = 1, n_virt_orb
aorb = list_virt(a)
lowest_super_ci_coef_mo(aorb,iorb) *= sqrt2
lowest_super_ci_coef_mo(iorb,aorb) *= sqrt2
enddo
enddo
! a_{ta} of the equation B.2 of Roos et. al. Chemical Physics 48 (1980) 157-173
do a = 1, n_virt_orb
aorb = list_virt(a)
do t = 1, n_act_orb
torb = list_act(t)
lowest_super_ci_coef_mo(torb,aorb) *= occ_act(t)**(-0.5d0)
lowest_super_ci_coef_mo(aorb,torb) *= occ_act(t)**(-0.5d0)
enddo
enddo
END_PROVIDER

View File

@ -1,132 +0,0 @@
BEGIN_PROVIDER [double precision, SXvector_lowest, (nMonoEx)]
implicit none
integer :: i
do i=2,nMonoEx+1
SXvector_lowest(i-1)=SXeigenvec(i,1)
enddo
END_PROVIDER
BEGIN_PROVIDER [double precision, thresh_overlap_switch]
implicit none
thresh_overlap_switch = 0.5d0
END_PROVIDER
BEGIN_PROVIDER [integer, max_overlap, (nMonoEx)]
&BEGIN_PROVIDER [integer, n_max_overlap]
&BEGIN_PROVIDER [integer, dim_n_max_overlap]
implicit none
double precision, allocatable :: vec_tmp(:)
integer, allocatable :: iorder(:)
allocate(vec_tmp(nMonoEx),iorder(nMonoEx))
integer :: i
do i = 1, nMonoEx
iorder(i) = i
vec_tmp(i) = -dabs(SXvector_lowest(i))
enddo
call dsort(vec_tmp,iorder,nMonoEx)
n_max_overlap = 0
do i = 1, nMonoEx
if(dabs(vec_tmp(i)).gt.thresh_overlap_switch)then
n_max_overlap += 1
max_overlap(n_max_overlap) = iorder(i)
endif
enddo
dim_n_max_overlap = max(1,n_max_overlap)
END_PROVIDER
BEGIN_PROVIDER [integer, orb_swap, (2,dim_n_max_overlap)]
&BEGIN_PROVIDER [integer, index_orb_swap, (dim_n_max_overlap)]
&BEGIN_PROVIDER [integer, n_orb_swap ]
implicit none
use bitmasks ! you need to include the bitmasks_module.f90 features
integer :: i,imono,iorb,jorb,j
n_orb_swap = 0
do i = 1, n_max_overlap
imono = max_overlap(i)
iorb = excit(1,imono)
jorb = excit(2,imono)
if (excit_class(imono) == "c-a" .and.hessmat2(imono,imono).gt.0.d0)then ! core --> active rotation
n_orb_swap += 1
orb_swap(1,n_orb_swap) = iorb ! core
orb_swap(2,n_orb_swap) = jorb ! active
index_orb_swap(n_orb_swap) = imono
else if (excit_class(imono) == "a-v" .and.hessmat2(imono,imono).gt.0.d0)then ! active --> virtual rotation
n_orb_swap += 1
orb_swap(1,n_orb_swap) = jorb ! virtual
orb_swap(2,n_orb_swap) = iorb ! active
index_orb_swap(n_orb_swap) = imono
endif
enddo
integer,allocatable :: orb_swap_tmp(:,:)
allocate(orb_swap_tmp(2,dim_n_max_overlap))
do i = 1, n_orb_swap
orb_swap_tmp(1,i) = orb_swap(1,i)
orb_swap_tmp(2,i) = orb_swap(2,i)
enddo
integer(bit_kind), allocatable :: det_i(:),det_j(:)
allocate(det_i(N_int),det_j(N_int))
logical, allocatable :: good_orb_rot(:)
allocate(good_orb_rot(n_orb_swap))
integer, allocatable :: index_orb_swap_tmp(:)
allocate(index_orb_swap_tmp(dim_n_max_overlap))
index_orb_swap_tmp = index_orb_swap
good_orb_rot = .True.
integer :: icount,k
do i = 1, n_orb_swap
if(.not.good_orb_rot(i))cycle
det_i = 0_bit_kind
call set_bit_to_integer(orb_swap(1,i),det_i,N_int)
call set_bit_to_integer(orb_swap(2,i),det_i,N_int)
do j = i+1, n_orb_swap
det_j = 0_bit_kind
call set_bit_to_integer(orb_swap(1,j),det_j,N_int)
call set_bit_to_integer(orb_swap(2,j),det_j,N_int)
icount = 0
do k = 1, N_int
icount += popcnt(ior(det_i(k),det_j(k)))
enddo
if (icount.ne.4)then
good_orb_rot(i) = .False.
good_orb_rot(j) = .False.
exit
endif
enddo
enddo
icount = n_orb_swap
n_orb_swap = 0
do i = 1, icount
if(good_orb_rot(i))then
n_orb_swap += 1
index_orb_swap(n_orb_swap) = index_orb_swap_tmp(i)
orb_swap(1,n_orb_swap) = orb_swap_tmp(1,i)
orb_swap(2,n_orb_swap) = orb_swap_tmp(2,i)
endif
enddo
if(n_orb_swap.gt.0)then
print*,'n_orb_swap = ',n_orb_swap
endif
do i = 1, n_orb_swap
print*,'imono = ',index_orb_swap(i)
print*,orb_swap(1,i),'-->',orb_swap(2,i)
enddo
END_PROVIDER
BEGIN_PROVIDER [double precision, switch_mo_coef, (ao_num,mo_num)]
implicit none
integer :: i,j,iorb,jorb
switch_mo_coef = NatOrbsFCI
do i = 1, n_orb_swap
iorb = orb_swap(1,i)
jorb = orb_swap(2,i)
do j = 1, ao_num
switch_mo_coef(j,jorb) = NatOrbsFCI(j,iorb)
enddo
do j = 1, ao_num
switch_mo_coef(j,iorb) = NatOrbsFCI(j,jorb)
enddo
enddo
END_PROVIDER

View File

@ -1,29 +0,0 @@
program test_pert_2rdm
implicit none
read_wf = .True.
touch read_wf
!call get_pert_2rdm
integer :: i,j,k,l,ii,jj,kk,ll
double precision :: accu , get_two_e_integral, integral
accu = 0.d0
print*,'n_orb_pert_rdm = ',n_orb_pert_rdm
do ii = 1, n_orb_pert_rdm
i = list_orb_pert_rdm(ii)
do jj = 1, n_orb_pert_rdm
j = list_orb_pert_rdm(jj)
do kk = 1, n_orb_pert_rdm
k= list_orb_pert_rdm(kk)
do ll = 1, n_orb_pert_rdm
l = list_orb_pert_rdm(ll)
integral = get_two_e_integral(i,j,k,l,mo_integrals_map)
! if(dabs(pert_2rdm_provider(ii,jj,kk,ll) * integral).gt.1.d-12)then
! print*,i,j,k,l
! print*,pert_2rdm_provider(ii,jj,kk,ll) * integral,pert_2rdm_provider(ii,jj,kk,ll), pert_2rdm_provider(ii,jj,kk,ll), integral
! endif
accu += pert_2rdm_provider(ii,jj,kk,ll) * integral
enddo
enddo
enddo
enddo
print*,'accu = ',accu
end

View File

@ -1,101 +0,0 @@
BEGIN_PROVIDER [real*8, etwo]
&BEGIN_PROVIDER [real*8, eone]
&BEGIN_PROVIDER [real*8, eone_bis]
&BEGIN_PROVIDER [real*8, etwo_bis]
&BEGIN_PROVIDER [real*8, etwo_ter]
&BEGIN_PROVIDER [real*8, ecore]
&BEGIN_PROVIDER [real*8, ecore_bis]
implicit none
integer :: t,u,v,x,i,ii,tt,uu,vv,xx,j,jj,t3,u3,v3,x3
real*8 :: e_one_all,e_two_all
e_one_all=0.D0
e_two_all=0.D0
do i=1,n_core_inact_orb
ii=list_core_inact(i)
e_one_all+=2.D0*mo_one_e_integrals(ii,ii)
do j=1,n_core_inact_orb
jj=list_core_inact(j)
e_two_all+=2.D0*bielec_PQxx(ii,ii,j,j)-bielec_PQxx(ii,jj,j,i)
end do
do t=1,n_act_orb
tt=list_act(t)
t3=t+n_core_inact_orb
do u=1,n_act_orb
uu=list_act(u)
u3=u+n_core_inact_orb
e_two_all+=D0tu(t,u)*(2.D0*bielec_PQxx(tt,uu,i,i) &
-bielec_PQxx(tt,ii,i,u3))
end do
end do
end do
do t=1,n_act_orb
tt=list_act(t)
do u=1,n_act_orb
uu=list_act(u)
e_one_all+=D0tu(t,u)*mo_one_e_integrals(tt,uu)
do v=1,n_act_orb
v3=v+n_core_inact_orb
do x=1,n_act_orb
x3=x+n_core_inact_orb
e_two_all +=P0tuvx(t,u,v,x)*bielec_PQxx(tt,uu,v3,x3)
end do
end do
end do
end do
ecore =nuclear_repulsion
ecore_bis=nuclear_repulsion
do i=1,n_core_inact_orb
ii=list_core_inact(i)
ecore +=2.D0*mo_one_e_integrals(ii,ii)
ecore_bis+=2.D0*mo_one_e_integrals(ii,ii)
do j=1,n_core_inact_orb
jj=list_core_inact(j)
ecore +=2.D0*bielec_PQxx(ii,ii,j,j)-bielec_PQxx(ii,jj,j,i)
ecore_bis+=2.D0*bielec_PxxQ(ii,i,j,jj)-bielec_PxxQ(ii,j,j,ii)
end do
end do
eone =0.D0
eone_bis=0.D0
etwo =0.D0
etwo_bis=0.D0
etwo_ter=0.D0
do t=1,n_act_orb
tt=list_act(t)
t3=t+n_core_inact_orb
do u=1,n_act_orb
uu=list_act(u)
u3=u+n_core_inact_orb
eone +=D0tu(t,u)*mo_one_e_integrals(tt,uu)
eone_bis+=D0tu(t,u)*mo_one_e_integrals(tt,uu)
do i=1,n_core_inact_orb
ii=list_core_inact(i)
eone +=D0tu(t,u)*(2.D0*bielec_PQxx(tt,uu,i,i) &
-bielec_PQxx(tt,ii,i,u3))
eone_bis+=D0tu(t,u)*(2.D0*bielec_PxxQ(tt,u3,i,ii) &
-bielec_PxxQ(tt,i,i,uu))
end do
do v=1,n_act_orb
vv=list_act(v)
v3=v+n_core_inact_orb
do x=1,n_act_orb
xx=list_act(x)
x3=x+n_core_inact_orb
real*8 :: h1,h2,h3
h1=bielec_PQxx(tt,uu,v3,x3)
h2=bielec_PxxQ(tt,u3,v3,xx)
h3=bielecCI(t,u,v,xx)
etwo +=P0tuvx(t,u,v,x)*h1
etwo_bis+=P0tuvx(t,u,v,x)*h2
etwo_ter+=P0tuvx(t,u,v,x)*h3
if ((h1.ne.h2).or.(h1.ne.h3)) then
write(6,9901) t,u,v,x,h1,h2,h3
9901 format('aie: ',4I4,3E20.12)
end if
end do
end do
end do
end do
END_PROVIDER

View File

@ -17,9 +17,12 @@ END_PROVIDER
pt2_F(:) = int(sqrt(float(pt2_n_tasks_max)))
do i=1,pt2_n_0(1+pt2_N_teeth/4)
pt2_F(i) = pt2_n_tasks_max
pt2_F(i) = pt2_n_tasks_max*pt2_min_parallel_tasks
enddo
do i=1+pt2_n_0(pt2_N_teeth-pt2_N_teeth/4), N_det_generators
do i=1+pt2_n_0(pt2_N_teeth-pt2_N_teeth/4), pt2_n_0(pt2_N_teeth-pt2_N_teeth/10)
pt2_F(i) = pt2_min_parallel_tasks
enddo
do i=1+pt2_n_0(pt2_N_teeth-pt2_N_teeth/10), N_det_generators
pt2_F(i) = 1
enddo

View File

@ -1,4 +1,3 @@
use bitmasks
BEGIN_PROVIDER [ double precision, pt2_match_weight, (N_states) ]
@ -144,7 +143,6 @@ BEGIN_PROVIDER [ double precision, selection_weight, (N_states) ]
END_PROVIDER
subroutine get_mask_phase(det1, pm, Nint)
use bitmasks
implicit none
@ -288,6 +286,7 @@ subroutine select_singles_and_doubles(i_generator,hole_mask,particle_mask,fock_d
PROVIDE psi_bilinear_matrix_rows psi_det_sorted_order psi_bilinear_matrix_order
PROVIDE psi_bilinear_matrix_transp_rows_loc psi_bilinear_matrix_transp_columns
PROVIDE psi_bilinear_matrix_transp_order psi_selectors_coef_transp
PROVIDE banned_excitation
monoAdo = .true.
monoBdo = .true.
@ -607,7 +606,8 @@ subroutine select_singles_and_doubles(i_generator,hole_mask,particle_mask,fock_d
h2 = hole_list(i2,s2)
call apply_hole(pmask, s2,h2, mask, ok, N_int)
banned = .false.
banned(:,:,1) = banned_excitation(:,:)
banned(:,:,2) = banned_excitation(:,:)
do j=1,mo_num
bannedOrb(j, 1) = .true.
bannedOrb(j, 2) = .true.
@ -674,7 +674,7 @@ subroutine fill_buffer_double(i_generator, sp, h1, h2, bannedOrb, banned, fock_d
logical :: ok
integer :: s1, s2, p1, p2, ib, j, istate, jstate
integer(bit_kind) :: mask(N_int, 2), det(N_int, 2)
double precision :: e_pert(N_states), coef(N_states), X(N_states)
double precision :: e_pert(N_states), coef(N_states)
double precision :: delta_E, val, Hii, w, tmp, alpha_h_psi
double precision, external :: diag_H_mat_elem_fock
double precision :: E_shift
@ -769,10 +769,16 @@ subroutine fill_buffer_double(i_generator, sp, h1, h2, bannedOrb, banned, fock_d
! call occ_pattern_of_det(det,occ,N_int)
! call occ_pattern_to_dets_size(occ,n,elec_alpha_num,N_int)
e_pert = 0.d0
coef = 0.d0
logical :: do_diag
do_diag = .False.
do istate=1,N_states
delta_E = E0(istate) - Hii + E_shift
alpha_h_psi = mat(istate, p1, p2)
if (alpha_h_psi == 0.d0) cycle
val = alpha_h_psi + alpha_h_psi
tmp = dsqrt(delta_E * delta_E + val * val)
if (delta_E < 0.d0) then
@ -784,13 +790,38 @@ subroutine fill_buffer_double(i_generator, sp, h1, h2, bannedOrb, banned, fock_d
else
coef(istate) = alpha_h_psi / delta_E
endif
if (e_pert(istate) < 0.d0) then
X(istate) = -dsqrt(-e_pert(istate))
else
X(istate) = dsqrt(e_pert(istate))
endif
enddo
do_diag = sum(dabs(coef)) > 0.001d0
double precision :: eigvalues(N_states+1)
double precision :: work(1+6*(N_states+1)+2*(N_states+1)**2)
integer :: iwork(3+5*(N_states+1)), info, k ,n
if (do_diag) then
double precision :: pt2_matrix(N_states+1,N_states+1)
pt2_matrix(N_states+1,N_states+1) = Hii+E_shift
do istate=1,N_states
pt2_matrix(:,istate) = 0.d0
pt2_matrix(istate,istate) = E0(istate)
pt2_matrix(istate,N_states+1) = mat(istate,p1,p2)
pt2_matrix(N_states+1,istate) = mat(istate,p1,p2)
enddo
call DSYEVD( 'V', 'U', N_states+1, pt2_matrix, N_states+1, eigvalues, &
work, size(work), iwork, size(iwork), info )
if (info /= 0) then
print *, 'error in '//irp_here
stop -1
endif
pt2_matrix = dabs(pt2_matrix)
iwork(1:N_states+1) = maxloc(pt2_matrix,DIM=1)
do k=1,N_states
e_pert(iwork(k)) = eigvalues(k) - E0(iwork(k))
enddo
endif
! ! Gram-Schmidt using input overlap matrix
! do istate=1,N_states
! do jstate=1,istate-1
@ -800,13 +831,12 @@ subroutine fill_buffer_double(i_generator, sp, h1, h2, bannedOrb, banned, fock_d
! enddo
do istate=1, N_states
alpha_h_psi = mat(istate, p1, p2)
do jstate=1,N_states
pt2_data % overlap(jstate,istate) += coef(jstate) * coef(istate)
enddo
enddo
do istate=1,N_states
alpha_h_psi = mat(istate, p1, p2)
pt2_data % variance(istate) += alpha_h_psi * alpha_h_psi
pt2_data % pt2(istate) += e_pert(istate)
@ -834,26 +864,29 @@ subroutine fill_buffer_double(i_generator, sp, h1, h2, bannedOrb, banned, fock_d
case(5)
! Variance selection
w = w - alpha_h_psi * alpha_h_psi * s_weight(istate,istate)
do jstate=1,N_states
if (istate == jstate) cycle
w = w + alpha_h_psi*mat(jstate,p1,p2) * s_weight(istate,jstate)
enddo
! w = w - alpha_h_psi * alpha_h_psi * s_weight(istate,istate)
w = min(w, - alpha_h_psi * alpha_h_psi * s_weight(istate,istate))
! do jstate=1,N_states
! if (istate == jstate) cycle
! w = w + dabs(alpha_h_psi*mat(jstate,p1,p2)) * s_weight(istate,jstate)
! enddo
case(6)
w = w - coef(istate) * coef(istate) * s_weight(istate,istate)
do jstate=1,N_states
if (istate == jstate) cycle
w = w + coef(istate)*coef(jstate) * s_weight(istate,jstate)
enddo
! w = w - coef(istate) * coef(istate) * s_weight(istate,istate)
w = min(w,- coef(istate) * coef(istate) * s_weight(istate,istate))
! do jstate=1,N_states
! if (istate == jstate) cycle
! w = w + dabs(coef(istate)*coef(jstate)) * s_weight(istate,jstate)
! enddo
case default
! Energy selection
w = w + e_pert(istate) * s_weight(istate,istate)
do jstate=1,N_states
if (istate == jstate) cycle
w = w - dabs(X(istate))*X(jstate) * s_weight(istate,jstate)
enddo
! w = w + e_pert(istate) * s_weight(istate,istate)
w = min(w, e_pert(istate) * s_weight(istate,istate))
! do jstate=1,N_states
! if (istate == jstate) cycle
! w = w + dabs(X(istate)*X(jstate)) * s_weight(istate,jstate)
! enddo
end select
end do

View File

@ -8,7 +8,7 @@ default: 1.e-10
type: logical
doc: Thresholds of Davidson's algorithm is set to E(rPT2)*threshold_davidson_from_pt2
interface: ezfio,provider,ocaml
default: true
default: false
[n_states_diag]
type: States_number

View File

@ -181,7 +181,7 @@ subroutine davidson_diag_hjj_sjj(dets_in,u_in,H_jj,s2_out,energies,dim_in,sze,N_
endif
overlap = 0.d0
PROVIDE nuclear_repulsion expected_s2 psi_bilinear_matrix_order psi_bilinear_matrix_order_reverse threshold_davidson_pt2
PROVIDE nuclear_repulsion expected_s2 psi_bilinear_matrix_order psi_bilinear_matrix_order_reverse threshold_davidson_pt2 threshold_davidson_from_pt2
call write_time(6)
write(6,'(A)') ''
@ -600,7 +600,7 @@ subroutine davidson_diag_hjj_sjj(dets_in,u_in,H_jj,s2_out,energies,dim_in,sze,N_
if ((itertot>1).and.(iter == 1)) then
!don't print
!don't print
continue
else
write(*,'(1X,I3,1X,100(1X,F16.10,1X,F11.6,1X,E11.3))') iter-1, to_print(1:3,1:N_st)
@ -608,9 +608,12 @@ subroutine davidson_diag_hjj_sjj(dets_in,u_in,H_jj,s2_out,energies,dim_in,sze,N_
! Check convergence
if (iter > 1) then
if (threshold_davidson_from_pt2) then
converged = dabs(maxval(residual_norm(1:N_st))) < threshold_davidson_pt2
endif
else
converged = dabs(maxval(residual_norm(1:N_st))) < threshold_davidson
endif
endif
do k=1,N_st
if (residual_norm(k) > 1.e8) then

View File

@ -59,7 +59,7 @@ function run_stoch() {
@test "HCO" { # 12.2868s
qp set_file hco.ezfio
run -113.389298880564 3.e-4 100000
run -113.389297812482 6.e-4 100000
}
@test "H2O2" { # 12.9214s
@ -175,7 +175,7 @@ function run_stoch() {
[[ -n $TRAVIS ]] && skip
qp set_file cu_nh3_4_2plus.ezfio
qp set_mo_class --core="[1-24]" --act="[25-45]" --del="[46-87]"
run -1862.98614665139 3.e-04 100000
run -1862.97568806589 3.e-04 100000
}
@test "HCN" { # 20.3273s

View File

@ -99,6 +99,10 @@ double precision function get_two_e_integral(i,j,k,l,map)
type(map_type), intent(inout) :: map
real(integral_kind) :: tmp
PROVIDE mo_two_e_integrals_in_map mo_integrals_cache
if (banned_excitation(i,k) .or. banned_excitation(j,l)) then
get_two_e_integral = 0.d0
return
endif
ii = l-mo_integrals_cache_min
ii = ior(ii, k-mo_integrals_cache_min)
ii = ior(ii, j-mo_integrals_cache_min)
@ -159,6 +163,11 @@ subroutine get_mo_two_e_integrals(j,k,l,sze,out_val,map)
! return
!DEBUG
out_val(1:sze) = 0.d0
if (banned_excitation(j,l)) then
return
endif
ii0 = l-mo_integrals_cache_min
ii0 = ior(ii0, k-mo_integrals_cache_min)
ii0 = ior(ii0, j-mo_integrals_cache_min)
@ -172,6 +181,7 @@ subroutine get_mo_two_e_integrals(j,k,l,sze,out_val,map)
q = q+shiftr(s*s-s,1)
do i=1,sze
if (banned_excitation(i,k)) cycle
ii = ior(ii0, i-mo_integrals_cache_min)
if (iand(ii, -128) == 0) then
ii_8 = ior( shiftl(ii0_8,7), int(i,8)-mo_integrals_cache_min_8)
@ -272,6 +282,29 @@ subroutine get_mo_two_e_integrals_exch_ii(k,l,sze,out_val,map)
end
BEGIN_PROVIDER [ logical, banned_excitation, (mo_num,mo_num) ]
implicit none
use map_module
BEGIN_DOC
! If true, the excitation is banned in the selection. Useful with local MOs.
END_DOC
banned_excitation = .False.
integer :: i,j
integer(key_kind) :: idx
double precision :: tmp
! double precision :: buffer(mo_num)
do j=1,mo_num
do i=1,j-1
call two_e_integrals_index(i,j,j,i,idx)
!DIR$ FORCEINLINE
call map_get(mo_integrals_map,idx,tmp)
banned_excitation(i,j) = dabs(tmp) < 1.d-15
banned_excitation(j,i) = banned_excitation(i,j)
enddo
enddo
END_PROVIDER
integer*8 function get_mo_map_size()
implicit none

View File

@ -16,6 +16,12 @@ doc: The selection process stops when the largest variance (for all the state) i
interface: ezfio,provider,ocaml
default: 0.0
[pt2_min_parallel_tasks]
type: integer
doc: Minimum number of tasks in PT2 calculation
interface: ezfio,provider,ocaml
default: 1
[pt2_relative_error]
type: Normalized_float
doc: Stop stochastic |PT2| when the relative error is smaller than `pT2_relative_error`

View File

@ -11,7 +11,7 @@ subroutine svd(A,LDA,U,LDU,D,Vt,LDVt,m,n)
integer, intent(in) :: LDA, LDU, LDVt, m, n
double precision, intent(in) :: A(LDA,n)
double precision, intent(out) :: U(LDU,m)
double precision, intent(out) :: U(LDU,min(m,n))
double precision,intent(out) :: Vt(LDVt,n)
double precision,intent(out) :: D(min(m,n))
double precision,allocatable :: work(:)
@ -19,19 +19,19 @@ subroutine svd(A,LDA,U,LDU,D,Vt,LDVt,m,n)
double precision,allocatable :: A_tmp(:,:)
allocate (A_tmp(LDA,n))
A_tmp = A
A_tmp(:,:) = A(:,:)
! Find optimal size for temp arrays
allocate(work(1))
lwork = -1
call dgesvd('A','A', m, n, A_tmp, LDA, &
call dgesvd('S','S', m, n, A_tmp, LDA, &
D, U, LDU, Vt, LDVt, work, lwork, info)
! /!\ int(WORK(1)) becomes negative when WORK(1) > 2147483648
lwork = max(int(work(1)), 5*MIN(M,N))
deallocate(work)
allocate(work(lwork))
call dgesvd('A','A', m, n, A_tmp, LDA, &
call dgesvd('S','S', m, n, A_tmp, LDA, &
D, U, LDU, Vt, LDVt, work, lwork, info)
deallocate(work,A_tmp)
@ -42,6 +42,128 @@ subroutine svd(A,LDA,U,LDU,D,Vt,LDVt,m,n)
end
subroutine eigSVD(A,LDA,U,LDU,D,Vt,LDVt,m,n)
implicit none
BEGIN_DOC
! Algorithm 3 of https://arxiv.org/pdf/1810.06860.pdf
!
! A(m,n) = U(m,n) D(n) Vt(n,n) with m>n
END_DOC
integer, intent(in) :: LDA, LDU, LDVt, m, n
double precision, intent(in) :: A(LDA,n)
double precision, intent(out) :: U(LDU,n)
double precision,intent(out) :: Vt(LDVt,n)
double precision,intent(out) :: D(n)
integer :: i,j,k
if (m<n) then
stop -1
call svd(A,LDA,U,LDU,D,Vt,LDVt,m,n)
return
endif
double precision, allocatable :: B(:,:), V(:,:)
allocate(B(n,n))
! B = - At . A
call dgemm('T','N',n,n,m,-1.d0,A,size(A,1),A,size(A,1),0.d0,B,size(B,1))
! V, D = eig(B)
allocate(V(n,n))
call lapack_diagd(D,V,B,n,n)
deallocate(B)
do j=1,n
do i=1,n
Vt(i,j) = V(j,i)
enddo
enddo
! S = sqrt(-D)
! U = A.V.S^-1
! U = A.(S^-1.vt)t
do k=1,n
if (D(k) >= 0.d0) then
exit
endif
D(k) = dsqrt(-D(k))
call dscal(n, 1.d0/D(k), V(1,k), 1)
enddo
D(k:n) = 0.d0
k=k-1
call dgemm('N','N',m,n,k,1.d0,A,size(A,1),V,size(V,1),0.d0,U,size(U,1))
end
subroutine randomized_svd(A,LDA,U,LDU,D,Vt,LDVt,m,n,q,r)
implicit none
include 'constants.include.F'
BEGIN_DOC
! Randomized SVD: rank r, q power iterations
!
! 1. Sample column space of A with P: Z = A.P where P is random with r+p columns.
!
! 2. Power iterations : Z <- X . (Xt.Z)
!
! 3. Z = Q.R
!
! 4. Compute SVD on projected Qt.X = U' . S. Vt
!
! 5. U = Q U'
END_DOC
integer, intent(in) :: LDA, LDU, LDVt, m, n, q, r
double precision, intent(in) :: A(LDA,n)
double precision, intent(out) :: U(LDU,r)
double precision,intent(out) :: Vt(LDVt,r)
double precision,intent(out) :: D(r)
integer :: i, j, k
double precision,allocatable :: Z(:,:), P(:,:), Y(:,:), UY(:,:)
double precision :: r1,r2
allocate(P(n,r), Z(m,r))
! P is a normal random matrix (n,r)
do k=1,r
do i=1,n
call random_number(r1)
call random_number(r2)
r1 = dsqrt(-2.d0*dlog(r1))
r2 = dtwo_pi*r2
P(i,k) = r1*dcos(r2)
enddo
enddo
! Z(m,r) = A(m,n).P(n,r)
call dgemm('N','N',m,r,n,1.d0,A,size(A,1),P,size(P,1),0.d0,Z,size(Z,1))
! Power iterations
do k=1,q
! P(n,r) = At(n,m).Z(m,r)
call dgemm('T','N',n,r,m,1.d0,A,size(A,1),Z,size(Z,1),0.d0,P,size(P,1))
! Z(m,r) = A(m,n).P(n,r)
call dgemm('N','N',m,r,n,1.d0,A,size(A,1),P,size(P,1),0.d0,Z,size(Z,1))
enddo
deallocate(P)
! QR factorization of Z
call ortho_svd(Z,size(Z,1),m,r)
allocate(Y(r,n), UY(r,r))
! Y(r,n) = Zt(r,m).A(m,n)
call dgemm('T','N',r,n,m,1.d0,Z,size(Z,1),A,size(A,1),0.d0,Y,size(Y,1))
! SVD of Y
call svd(Y,size(Y,1),UY,size(UY,1),D,Vt,size(Vt,1),r,n)
deallocate(Y)
! U(m,r) = Z(m,r).UY(r,r)
call dgemm('N','N',m,r,r,1.d0,Z,size(Z,1),UY,size(UY,1),0.d0,U,size(U,1))
deallocate(UY,Z)
end
subroutine svd_complex(A,LDA,U,LDU,D,Vt,LDVt,m,n)
implicit none
@ -807,6 +929,33 @@ subroutine ortho_canonical(overlap,LDA,N,C,LDC,m,cutoff)
end
subroutine ortho_svd(A,LDA,m,n)
implicit none
BEGIN_DOC
! Orthogonalization via fast SVD
!
! A : matrix to orthogonalize
!
! LDA : leftmost dimension of A
!
! m : Number of rows of A
!
! n : Number of columns of A
!
END_DOC
integer, intent(in) :: m,n, LDA
double precision, intent(inout) :: A(LDA,n)
if (m < n) then
call ortho_qr(A,LDA,m,n)
endif
double precision, allocatable :: U(:,:), D(:), Vt(:,:)
allocate(U(m,n), D(n), Vt(n,n))
call SVD(A,LDA,U,size(U,1),D,Vt,size(Vt,1),m,n)
A(1:m,1:n) = U(1:m,1:n)
deallocate(U,D, Vt)
end
subroutine ortho_qr(A,LDA,m,n)
implicit none
BEGIN_DOC