Acceleration of filter_connected and Davidson

2024-12-22 20:35:19 +01:00 · 2014-05-15 17:58:30 +02:00 · 2014-05-15 17:58:30 +02:00 · 13ee3531ee
commit 13ee3531ee
parent df9efae335
11 changed files with 108 additions and 55 deletions
--- a/src/AOs/tests/Makefile
+++ b/src/AOs/tests/Makefile
@ -0,0 +1,33 @@
+OPENMP  =1
+PROFILE =0
+DEBUG  = 0
+
+IRPF90+= -I tests
+
+REF_FILES=$(subst %.irp.f, %.ref, $(wildcard *.irp.f))
+
+.PHONY: clean executables serial_tests parallel_tests
+
+all: clean executables serial_tests parallel_tests
+
+parallel_tests: $(REF_FILES)
+	@echo ; echo "   ----  Running parallel tests ----" ; echo
+	@OMP_NUM_THREADS=10 ${QPACKAGE_ROOT}/scripts/run_tests.py
+
+serial_tests: $(REF_FILES)
+	@echo ; echo "   ----  Running serial tests ----" ; echo
+	@OMP_NUM_THREADS=1 ${QPACKAGE_ROOT}/scripts/run_tests.py
+
+executables: $(wildcard *.irp.f) veryclean
+	$(MAKE) -C ..
+
+%.ref: $(wildcard $(QPACKAGE_ROOT)/data/inputs/*.md5) executables
+	$(QPACKAGE_ROOT)/scripts/create_test_ref.sh $*
+
+clean:
+	$(MAKE) -C .. clean
+
+veryclean:
+	$(MAKE) -C .. veryclean
+
+
--- a/src/BiInts/README.rst
+++ b/src/BiInts/README.rst
@ -107,7 +107,8 @@ Documentation
  AO integrals

 `bielec_integrals_index <http://github.com/LCPQ/quantum_package/tree/master/src/BiInts/map_integrals.irp.f#L17>`_
-None
+  Undocumented
+
 `clear_ao_map <http://github.com/LCPQ/quantum_package/tree/master/src/BiInts/map_integrals.irp.f#L128>`_
  Frees the memory of the AO map

--- a/src/CISD/cisd.irp.f
+++ b/src/CISD/cisd.irp.f
@ -1,11 +1,16 @@
 program cisd
  implicit none
-  integer :: i
+  integer :: i,k
  double precision, allocatable  :: eigvalues(:),eigvectors(:,:)
+  PROVIDE ref_bitmask_energy
  call H_apply_cisd
-  allocate(eigvalues(n_det),eigvectors(n_det,n_det))
+  allocate(eigvalues(n_states),eigvectors(n_det,n_states))
  print *,  'N_det = ', N_det
-  psi_coef = psi_coef - 1.d-4
+  print *,  'N_states = ', N_states
+  psi_coef = - 1.d-4
+  do k=1,N_states
+    psi_coef(k,k) = 1.d0
+  enddo
  call davidson_diag(psi_det,psi_coef,eigvalues,size(psi_coef,1),N_det,N_states,N_int)

  print *,  '---'
--- a/src/Dets/H_apply_template.f
+++ b/src/Dets/H_apply_template.f
@ -26,7 +26,7 @@ subroutine $subroutine_diexc(key_in, hole_1,particl_1, hole_2, particl_2 $parame
  integer, allocatable           :: ia_ja_pairs(:,:,:)
  double precision               :: diag_H_mat_elem, E_ref
  
-  PROVIDE mo_integrals_map
+  PROVIDE mo_integrals_map ref_bitmask_energy
  PROVIDE mo_bielec_integrals_in_map
  
  $set_i_H_j_threshold
@ -257,7 +257,7 @@ subroutine $subroutine_monoexc(key_in, hole_1,particl_1 $parameters )
  integer, allocatable           :: ia_ja_pairs(:,:,:)
  double precision               :: diag_H_mat_elem, E_ref
  
-  PROVIDE mo_integrals_map
+  PROVIDE mo_integrals_map ref_bitmask_energy
  PROVIDE mo_bielec_integrals_in_map

  $set_i_H_j_threshold
--- a/src/Dets/davidson.irp.f
+++ b/src/Dets/davidson.irp.f
@ -12,7 +12,7 @@ BEGIN_PROVIDER [ integer, davidson_sze_max]
  ! Max number of Davidson sizes
  END_DOC
  ASSERT (davidson_sze_max <= davidson_iter_max)
-  davidson_sze_max = 4
+  davidson_sze_max = 8
 END_PROVIDER

 subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
@ -50,15 +50,17 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
  integer                        :: k_pairs, kl
  
  integer                        :: iter2
-  double precision, allocatable  :: W(:,:), H_jj(:), U(:,:,:), R(:,:)
+  double precision, allocatable  :: W(:,:,:), H_jj(:), U(:,:,:), R(:,:)
  double precision, allocatable  :: y(:,:,:,:), h(:,:,:,:), lambda(:)
  double precision               :: diag_h_mat_elem
  double precision               :: residual_norm(N_st)
  
+  PROVIDE ref_bitmask_energy
+
  allocate(                                                          &
      kl_pairs(2,N_st*(N_st+1)/2),                                   &
      H_jj(sze),                                                     &
-      W(sze,N_st),                                                   &
+      W(sze,N_st,davidson_sze_max),                                                   &
      U(sze,N_st,davidson_sze_max),                                  &
      R(sze,N_st),                                                   &
      h(N_st,davidson_sze_max,N_st,davidson_sze_max),                &
@ -109,6 +111,7 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
  enddo
  !$OMP END DO
  !$OMP END PARALLEL
+
  call ortho_lowdin(overlap,size(overlap,1),N_st,U_in,size(U_in,1),sze)
  
  ! Davidson iterations
@ -148,33 +151,24 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
      ! ----------------------
      
      do k=1,N_st
-        call H_u_0(W(1,k),U(1,k,iter),H_jj,sze,dets_in,Nint)
+        call H_u_0(W(1,k,iter),U(1,k,iter),H_jj,sze,dets_in,Nint)
      enddo
      
      ! Compute h_kl = <u_k | W_l> = <u_k| H |u_l>
      ! -------------------------------------------

-      !$OMP PARALLEL 
-      !$OMP SINGLE
      do l=1,N_st
-        do iter2=1,iter-1
-          do k=1,N_st
-            !$OMP TASK FIRSTPRIVATE(k,iter,l,iter2)
-            h(k,iter2,l,iter) = u_dot_v(U(1,k,iter2),W(1,l),sze)
+        do k=1,N_st
+          do iter2=1,iter-1
+            h(k,iter2,l,iter) = u_dot_v(U(1,k,iter2),W(1,l,iter),sze)
            h(k,iter,l,iter2) = h(k,iter2,l,iter)
-            !$OMP END TASK
          enddo
        enddo
        do k=1,l
-          !$OMP TASK FIRSTPRIVATE(k,iter,l)
-          h(k,iter,l,iter) = u_dot_v(U(1,k,iter),W(1,l),sze)
+          h(k,iter,l,iter) = u_dot_v(U(1,k,iter),W(1,l,iter),sze)
          h(l,iter,k,iter) = h(k,iter,l,iter)
-          !$OMP END TASK
        enddo
      enddo
-      !$OMP END SINGLE NOWAIT
-      !$OMP TASKWAIT
-      !$OMP END PARALLEL
      
      ! Diagonalize h
      ! -------------
@ -184,13 +178,17 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
      ! Express eigenvectors of h in the determinant basis
      ! --------------------------------------------------
      
-      !TODO dgemm
+  !   call dgemm ( 'N','N', sze, N_st*iter, N_st, &
+  !     1.d0, U(1,1,1), size(U,1), y(1,1,1,1), size(y,1)*size(y,2), &
+  !     0.d0, U(1,1,iter+1), size(U,1) )
      do k=1,N_st
        do i=1,sze
          U(i,k,iter+1) = 0.d0
-          do iter2=1,iter
-            do l=1,N_st
-              U(i,k,iter+1) += U(i,l,iter2)*y(l,iter2,k,1)
+          W(i,k,iter+1) = 0.d0
+          do l=1,N_st
+            do iter2=1,iter
+              U(i,k,iter+1) = U(i,k,iter+1) + U(i,l,iter2)*y(l,iter2,k,1)
+              W(i,k,iter+1) = W(i,k,iter+1) + W(i,l,iter2)*y(l,iter2,k,1)
            enddo
          enddo
        enddo
@ -199,13 +197,9 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
      ! Compute residual vector
      ! -----------------------
      
-      do k=1,N_st
-        call H_u_0(W(1,k),U(1,k,iter+1),H_jj,sze,dets_in,Nint)
-      enddo
-      
      do k=1,N_st
        do i=1,sze
-          R(i,k) = lambda(k) * U(i,k,iter+1) - W(i,k)
+          R(i,k) = lambda(k) * U(i,k,iter+1) - W(i,k,iter+1)
        enddo
        residual_norm(k) = u_dot_u(R(1,k),sze)
      enddo
@ -215,7 +209,7 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
      print *,  residual_norm(1:N_st)
      print *,  ''
      
-      converged = maxval(residual_norm) < 1.d-10
+      converged = maxval(residual_norm) < 1.d-5
      if (converged) then
        exit
      endif
@ -253,7 +247,7 @@ subroutine davidson_diag(dets_in,u_in,energies,dim_in,sze,N_st,Nint)
    enddo
    
    if (.not.converged) then
-      iter = davidson_sze_max
+      iter = davidson_sze_max-1
    endif
    
    ! Re-contract to u_in
--- a/src/Dets/slater_rules.irp.f
+++ b/src/Dets/slater_rules.irp.f
@ -964,26 +964,38 @@ subroutine H_u_0(v_0,u_0,H_jj,n,keys_tmp,Nint)
  integer, allocatable           :: idx(:)
  double precision               :: hij
  integer                        :: i,j,k,l, jj
+  integer                        :: i0, j0
  ASSERT (Nint > 0)
  ASSERT (Nint == N_int)
-  ASSERT (n>0)
+  ASSERT (n>0) 
+  PROVIDE ref_bitmask_energy
+  integer, parameter :: block_size = 157
  !$OMP PARALLEL DEFAULT(NONE)                                       &
      !$OMP PRIVATE(i,hij,j,k,idx,jj) SHARED(n,H_jj,u_0,keys_tmp,Nint)&
      !$OMP SHARED(v_0)
-  allocate(idx(0:n))
-  !$OMP DO SCHEDULE(guided)
+  allocate(idx(0:block_size))
+  !$OMP DO SCHEDULE(static)
  do i=1,n
    v_0(i) = H_jj(i) * u_0(i)
-    call filter_connected(keys_tmp,keys_tmp(1,1,i),Nint,n,idx)
-    do jj=1,idx(0)
-      j = idx(jj)
-      if (j/=i) then
-        call i_H_j(keys_tmp(1,1,j),keys_tmp(1,1,i),Nint,hij)
-        v_0(i) = v_0(i) + hij*u_0(j)
-      endif
-    enddo
  enddo
  !$OMP END DO
+  !$OMP DO SCHEDULE(guided)
+  do i0=1,n,block_size
+    do j0=1,n,block_size
+      do i=i0,min(i0+block_size-1,n)
+        call filter_connected(keys_tmp(1,1,j0),keys_tmp(1,1,i),Nint,min(block_size,i-j0+1),idx)
+        do jj=1,idx(0)
+          j = idx(jj)+j0-1
+          if ( (j<i).and.(dabs(u_0(j)) > 1.d-8)) then
+            call i_H_j(keys_tmp(1,1,j),keys_tmp(1,1,i),Nint,hij)
+            v_0(i) = v_0(i) + hij*u_0(j)
+            v_0(j) = v_0(j) + hij*u_0(i)
+          endif
+        enddo
+      enddo
+    enddo
+  enddo
+  !$OMP END DO NOWAIT
  deallocate(idx)
  !$OMP END PARALLEL
 end
--- a/src/Hartree_Fock/README.rst
+++ b/src/Hartree_Fock/README.rst
@ -82,7 +82,8 @@ Documentation
  Diagonal Fock matrix in the MO basis

 `scf_iteration <http://github.com/LCPQ/quantum_package/tree/master/src/Hartree_Fock/mo_SCF_iterations.irp.f#L1>`_
-None
+  Undocumented
+
 `do_diis <http://github.com/LCPQ/quantum_package/tree/master/src/Hartree_Fock/options.irp.f#L41>`_
  If True, compute integrals on the fly

--- a/src/MOs/README.rst
+++ b/src/MOs/README.rst
@ -54,8 +54,10 @@ Documentation
  Aligned variable for dimensioning of arrays

 `mo_as_eigvectors_of_mo_matrix <http://github.com/LCPQ/quantum_package/tree/master/src/MOs/utils.irp.f#L21>`_
-None
+  Undocumented
+
 `save_mos <http://github.com/LCPQ/quantum_package/tree/master/src/MOs/utils.irp.f#L1>`_
-None
+  Undocumented
+


--- a/src/MonoInts/kin_ao_ints.irp.f
+++ b/src/MonoInts/kin_ao_ints.irp.f
@ -11,6 +11,7 @@
  double precision :: A_center(3), B_center(3)
  integer :: power_A(3), power_B(3)
  double precision :: d_a_2,d_2
+  PROVIDE all_utils 
  dim1=100
  BEGIN_DOC
  ! second derivatives matrix elements in the ao basis 
--- a/src/Utils/map_module.f90
+++ b/src/Utils/map_module.f90
@ -482,7 +482,7 @@ subroutine cache_map_get_interval(map, key, value, ibegin, iend, idx)
 if (idx > 0) then
   value = map%value(idx)
 else
-   value = 0.
+   value = 0._integral_kind
 endif
 end

--- a/src/Utils/util.irp.f
+++ b/src/Utils/util.irp.f
@ -253,12 +253,16 @@ double precision function u_dot_u(u,sze)
  t3 = t2+t2
  t4 = t3+t2
  u_dot_u = 0.d0
-  do i=1,t2
-    u_dot_u = u_dot_u + u(t1+i)*u(t1+i) + u(t2+i)*u(t2+i) +          &
-        u(t3+i)*u(t3+i) + u(t4+i)*u(t4+i)
-  enddo
-  do i=t4+t2+1,sze
-    u_dot_u = u_dot_u+u(i)*u(i)
+! do i=1,t2
+!   u_dot_u = u_dot_u + u(t1+i)*u(t1+i) + u(t2+i)*u(t2+i) +          &
+!       u(t3+i)*u(t3+i) + u(t4+i)*u(t4+i)
+! enddo
+! do i=t4+t2+1,sze
+!   u_dot_u = u_dot_u+u(i)*u(i)
+! enddo
+  
+  do i=1,sze
+    u_dot_u = u_dot_u + u(i)*u(i)
  enddo
  
 end