TREX/irpjast
1
0
Fork 0
mirror of https://github.com/TREX-CoE/irpjast.git synced 2024-12-22 12:23:57 +01:00
Code Issues Projects Releases Wiki Activity

Merge pull request #3 from Panadestein/as

Browse Source 
BLAS jastrow (value only)
This commit is contained in:
Ramon L. PANADES-BARRUETA 2021-01-19 15:04:20 +01:00 committed by GitHub
commit 71a78333d8
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
7 changed files with 241 additions and 51 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Makefile
View File

@ -1,5 +1,5 @@
IRPF90 = irpf90 --codelet=factor_een:100000 #-a -d
FC = ifort -xHost -g
FC = ifort -xHost -g -mkl=sequential
FCFLAGS= -O2 -ffree-line-length-none -I .
NINJA = ninja
AR = ar

34
README.org
View File

@ -1,3 +1,37 @@
* IRPJAST
CHAMP's Jastrow factor computation using the IRPF90 method
Original equation:
$$
\sum_{i=2}^{Ne} \sum_{j=1}^i \sum_{pkl} \sum_a^{Nn} c_{apkl}\, r_{ij}^k\, ( R_{ia}^l + R_{ja}^l) ( R_{ia} R_{ja})^m
$$
Expanding, one obtains:
$$
\sum_{i=2}^{Ne} \sum_{j=1}^i \sum_{pkl} \sum_a^{Nn} c_{apkl} R_{ia}^{p-k-l}\, r_{ij}^k\, R_{ja}^{p-k+l} + c_{apkl} R_{ia}^{p-k+l}\, r_{ij}^k\, R_{ja}^{p-k-l}
$$
The equation is symmetric if we exchange $i$ and $j$, so we can rewrite
$$
\sum_{i=1}^{Ne} \sum_{j=1}^{Ne} \sum_{pkl} \sum_a^{Nn} c_{apkl} R_{ia}^{p-k-l}\, r_{ij}^k\, R_{ja}^{p-k+l}
$$
If we reshape $R_{ja}^p$ as a matrix $R_{j,al}$ of size
$N_e \times N_n(N_c+1)$,
for every $k$, we can pre-compute the matrix product
$$
C_{i,al}^{k} = \sum_j r_{ij}^k\, R_{i,al}
$$
which can be computed efficiently with BLAS.
We can express the total Jastrow as:
$$
\sum_{i=1}^{Ne} \sum_{pkl} \sum_a^{Nn}
c_{apkl} R_{ia}^{p-k-l}\, C_{i,a(p-k+l)}^k
$$

BIN
deriv_num
View File

Binary file not shown.

47
el_nuc_el.irp.f
View File

@ -2,10 +2,16 @@ BEGIN_PROVIDER [ double precision, factor_een ]
implicit none
BEGIN_DOC
! ElectronE-electron-nuclei contribution to Jastrow factor
!
! 5436.20340250000
END_DOC
integer :: i, j, a, p, k, l, lmax, m
double precision :: rjam_cn
double precision :: cn
integer :: i, j, a, p, k, l, lmax, m, n
double precision :: cn, accu2, accu
double precision :: f(nnuc,0:ncord-2,0:ncord-2)
double precision :: tmp_c(nelec,nnuc,0:ncord,0:ncord-1)
! factor_een = factor_een_blas
! return
factor_een = 0.0d0
@ -18,22 +24,25 @@ BEGIN_PROVIDER [ double precision, factor_een ]
endif
do l = 0, lmax
if ( iand(p - k - l, 1) == 1) cycle
m = (p - k - l) / 2
do a = 1, nnuc
accu2 = 0.d0
cn = cord_vect_lkp(l, k, p, typenuc_arr(a))
rjam_cn = rescale_een_n(2, a, m) * cn
factor_een = factor_een + rescale_een_e(1,2,k) * &
(rescale_een_n(1,a,l) + rescale_een_n(2,a,l)) * &
rescale_een_n(1,a,m) * rjam_cn
do j = 3, nelec
rjam_cn = rescale_een_n(j, a, m) * cn
do i = 1, j - 1
factor_een = factor_een + rescale_een_e(i,j,k) * &
(rescale_een_n(i,a,l) + rescale_een_n(j,a,l)) * &
rescale_een_n(i,a,m) * rjam_cn
do j = 1, nelec
accu = 0.d0
do i = 1, nelec
accu = accu + &
rescale_een_e(i,j,k) * &
rescale_een_n(i,a,m) * &
rescale_een_n(j,a,m+l)
enddo
accu2 = accu2 + accu
enddo
factor_een = factor_een + accu2 * cn
enddo
enddo
enddo
enddo
@ -78,7 +87,7 @@ BEGIN_PROVIDER [ double precision, factor_een_deriv_e, (4, nelec) ]
enddo
do i = 1, nelec
rial = rescale_een_n(i, a, l)
rial = rescale_een_n(i, a, l) + rjal
riam = rescale_een_n(i, a, m)
rijk = rescale_een_e(i, j, k)
@ -86,24 +95,24 @@ BEGIN_PROVIDER [ double precision, factor_een_deriv_e, (4, nelec) ]
drijk(ii) = rescale_een_e_deriv_e(ii, j, i, k)
enddo
v1 = rijk * (rial + rjal) ! v(x)
v1 = rijk * rial ! v(x)
v2 = rjam_cn * riam ! u(x)
lap1 = 0.0d0
lap2 = 0.0d0
do ii = 1, 3
d1 = drijk(ii) * (rial + rjal) + rijk * drjal(ii)
d1 = drijk(ii) * rial + rijk * drjal(ii)
d2 = drjam_cn(ii) * riam
lap1 = lap1 + d1 * d2
lap2 = lap2 + drijk(ii) * drjal(ii)
factor_een_deriv_e(ii, j) += v1 * d2 + d1 * v2
factor_een_deriv_e(ii, j) = factor_een_deriv_e(ii, j) + v1 * d2 + d1 * v2
enddo
! v(x) u''(x) + 2 * u'(x) v'(x) + u(x) v''(x)
ii = 4
d1 = drijk(ii) * (rial + rjal) + rijk * drjal(ii) + 2.0d0 * lap2
d1 = drijk(ii) * rial + rijk * drjal(ii) + lap2 + lap2
d2 = drjam_cn(ii) * riam
factor_een_deriv_e(ii, j) += v1 * d2 + d1 * v2 + 2.0d0 * lap1
factor_een_deriv_e(ii, j) = factor_een_deriv_e(ii, j) + v1 * d2 + d1 * v2 + lap1 + lap1
enddo
enddo

126
el_nuc_el_blas.irp.f Normal file
View File

@ -0,0 +1,126 @@
BEGIN_PROVIDER [ double precision, factor_een_blas ]
implicit none
BEGIN_DOC
! ElectronE-electron-nuclei contribution to Jastrow factor
!
! 4124.84239750000
END_DOC
integer :: i, j, a, p, k, l, lmax, m, n
double precision :: cn(nnuc), accu
double precision :: f(nnuc,0:ncord-2,0:ncord-2)
double precision :: tmp_c(nelec,nnuc,0:ncord,0:ncord-1)
factor_een_blas = 0.0d0
! r_{ij}^k . R_{ja}^l -> tmp_c_{ia}^{kl}
do k=0,ncord-1
call dgemm('N','N', nelec, nnuc*(ncord+1), nelec, 1.d0, &
rescale_een_e(1,1,k), size(rescale_een_e,1), &
rescale_een_n(1,1,0), size(rescale_een_n,1), 0.d0, &
tmp_c(1,1,0,k), size(tmp_c,1))
enddo
do p = 2, ncord
do k = 0, p - 1
m = p-k
if (k > 0) then
lmax = m
else
lmax = m - 2
endif
n = shiftr(m,1)
do l = iand(m, 1), lmax, 2
do a = 1, nnuc
cn(a) = cord_vect_lkp(l, k, p, typenuc_arr(a))
enddo
do a = 1, nnuc
accu = 0.d0
do i=1,nelec
accu = accu + rescale_een_n(i,a,n) * tmp_c(i,a,n+l,k)
enddo
factor_een_blas = factor_een_blas + accu * cn(a)
enddo
n = n-1
enddo
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [ double precision, factor_een_deriv_e_blas, (4, nelec) ]
implicit none
BEGIN_DOC
! Dimensions 1-3 : dx, dy, dz
! Dimension 4 : d2x + d2y + d2z
END_DOC
integer :: i, ii, j, a, p, k, l, lmax, m
double precision :: riam, rjam_cn, rial, rjal, rijk
double precision, dimension(4) :: driam, drjam_cn, drial, drjal, drijk
double precision :: cn, v1, v2, d1, d2, lap1, lap2
factor_een_deriv_e_blas = 0.0d0
do p = 2, ncord
do k = 0 , p - 1
if (k /= 0) then
lmax = p - k
else
lmax = p - k - 2
endif
do l = 0, lmax
if ( iand(p - k - l, 1) == 1) cycle
m = (p - k - l) / 2
do a = 1, nnuc
cn = cord_vect_lkp(l, k, p, typenuc_arr(a))
do j = 1, nelec
rjal = rescale_een_n(j, a, l)
rjam_cn = rescale_een_n(j, a, m) * cn
do ii = 1, 4
drjal(ii) = rescale_een_n_deriv_e(ii, j, a, l)
drjam_cn(ii) = rescale_een_n_deriv_e(ii, j, a, m) * cn
enddo
do i = 1, nelec
rial = rescale_een_n(i, a, l) + rjal
riam = rescale_een_n(i, a, m)
rijk = rescale_een_e(i, j, k)
do ii = 1, 4
drijk(ii) = rescale_een_e_deriv_e(ii, j, i, k)
enddo
v1 = rijk * rial ! v(x)
v2 = rjam_cn * riam ! u(x)
lap1 = 0.0d0
lap2 = 0.0d0
do ii = 1, 3
d1 = drijk(ii) * rial + rijk * drjal(ii)
d2 = drjam_cn(ii) * riam
lap1 = lap1 + d1 * d2
lap2 = lap2 + drijk(ii) * drjal(ii)
factor_een_deriv_e_blas(ii, j) = factor_een_deriv_e_blas(ii, j) + v1 * d2 + d1 * v2
enddo
! v(x) u''(x) + 2 * u'(x) v'(x) + u(x) v''(x)
ii = 4
d1 = drijk(ii) * rial + rijk * drjal(ii) + lap2 + lap2
d2 = drjam_cn(ii) * riam
factor_een_deriv_e_blas(ii, j) = factor_een_deriv_e_blas(ii, j) + v1 * d2 + d1 * v2 + lap1 + lap1
enddo
enddo
enddo
enddo
enddo
enddo
END_PROVIDER

BIN
jastrow
View File

Binary file not shown.

43
rescale.irp.f
View File

@ -107,24 +107,18 @@ BEGIN_PROVIDER [double precision, rescale_een_e, (nelec, nelec, 0:ncord)]
BEGIN_DOC
! R = exp(-kappa r) for electron-electron for $J_{een}$
END_DOC
integer :: i, j, l
integer :: i, j, k, l
double precision :: x
double precision, parameter :: f = dexp(1.d0)
rescale_een_e(:, :, 0) = 1.d0
do l = 1, ncord
k=0
do j = 1, nelec
do i = 1, j-1
x = dexp(-kappa * elec_dist(i, j))
rescale_een_e(i, j, 1) = x
rescale_een_e(j, i, 1) = x
enddo
enddo
do l = 2, ncord
do j = 1, nelec
do i = 1, j-1
x = rescale_een_e(i, j, l-1) * rescale_een_e(i, j, 1)
k = k+1
x = rescale_een_e_ij(k,l)
rescale_een_e(i, j, l) = x
rescale_een_e(j, i, l) = x
enddo
@ -138,6 +132,33 @@ BEGIN_PROVIDER [double precision, rescale_een_e, (nelec, nelec, 0:ncord)]
enddo
END_PROVIDER
BEGIN_PROVIDER [double precision, rescale_een_e_ij, (nelec*(nelec-1)/2, 0:ncord)]
implicit none
BEGIN_DOC
! R = exp(-kappa r) for electron-electron for $J_{een}$
END_DOC
integer :: i, j, l,k
double precision :: x
double precision, parameter :: f = dexp(1.d0)
rescale_een_e_ij(:, 0) = 1.d0
k=0
do j = 1, nelec
do i = 1, j-1
k = k+1
rescale_een_e_ij(k, 1) = dexp(-kappa * elec_dist(i, j))
enddo
enddo
do l = 2, ncord
do k=1,(nelec*nelec-nelec)/2
rescale_een_e_ij(k, l) = rescale_een_e_ij(k, l-1) * rescale_een_e_ij(k, 1)
enddo
enddo
END_PROVIDER
BEGIN_PROVIDER [double precision, rescale_een_n, (nelec, nnuc, 0:ncord)]
implicit none
BEGIN_DOC