9
1
mirror of https://github.com/QuantumPackage/qp2.git synced 2024-12-22 11:33:29 +01:00
qp2/plugins/local/ao_tc_eff_map/fit_j.irp.f

520 lines
15 KiB
Fortran
Raw Normal View History

2023-02-06 19:00:35 +01:00
BEGIN_PROVIDER [ double precision, expo_j_xmu_1gauss ]
&BEGIN_PROVIDER [ double precision, coef_j_xmu_1gauss ]
implicit none
BEGIN_DOC
! Upper bound long range fit of F(x) = x * (1 - erf(x)) - 1/sqrt(pi) * exp(-x**2)
!
! with a single gaussian.
!
! Such a function can be used to screen integrals with F(x).
END_DOC
expo_j_xmu_1gauss = 0.5d0
coef_j_xmu_1gauss = 1.d0
END_PROVIDER
! ---
BEGIN_PROVIDER [ double precision, expo_erfc_gauss ]
implicit none
expo_erfc_gauss = 1.41211d0
END_PROVIDER
BEGIN_PROVIDER [ double precision, expo_erfc_mu_gauss ]
implicit none
expo_erfc_mu_gauss = expo_erfc_gauss * mu_erf * mu_erf
END_PROVIDER
BEGIN_PROVIDER [ double precision, expo_good_j_mu_1gauss ]
&BEGIN_PROVIDER [ double precision, coef_good_j_mu_1gauss ]
implicit none
BEGIN_DOC
! exponent of Gaussian in order to obtain an upper bound of J(r12,mu)
!
! Can be used to scree integrals with J(r12,mu)
END_DOC
expo_good_j_mu_1gauss = 2.D0 * mu_erf * expo_j_xmu_1gauss
coef_good_j_mu_1gauss = 0.5d0/mu_erf * coef_j_xmu_1gauss
END_PROVIDER
BEGIN_PROVIDER [ double precision, expo_j_xmu, (n_fit_1_erf_x) ]
BEGIN_DOC
! F(x) = x * (1 - erf(x)) - 1/sqrt(pi) * exp(-x**2) is fitted with a gaussian and a Slater
!
! \approx - 1/sqrt(pi) * exp(-alpha * x ) exp(-beta * x**2)
!
! where alpha = expo_j_xmu(1) and beta = expo_j_xmu(2)
END_DOC
implicit none
!expo_j_xmu(1) = 1.7477d0
!expo_j_xmu(2) = 0.668662d0
!expo_j_xmu(1) = 1.74766377595541d0
!expo_j_xmu(2) = 0.668719925486403d0
expo_j_xmu(1) = 1.74770446934522d0
expo_j_xmu(2) = 0.668659706559979d0
2023-02-06 19:00:35 +01:00
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, expo_gauss_j_mu_x, (ng_fit_jast)]
&BEGIN_PROVIDER [double precision, coef_gauss_j_mu_x, (ng_fit_jast)]
BEGIN_DOC
!
! J(mu,r12) = 1/2 r12 * (1 - erf(mu*r12)) - 1/(2 sqrt(pi)*mu) exp(-(mu*r12)^2) is expressed as
!
! J(mu,r12) = 0.5/mu * F(r12*mu) where F(x) = x * (1 - erf(x)) - 1/sqrt(pi) * exp(-x**2)
!
! F(x) is fitted by - 1/sqrt(pi) * exp(-alpha * x) exp(-beta * x^2) (see expo_j_xmu)
!
! The slater function exp(-alpha * x) is fitted with n_max_fit_slat gaussians
!
! See Appendix 2 of JCP 154, 084119 (2021)
!
END_DOC
implicit none
integer :: i
double precision :: tmp
double precision :: expos(ng_fit_jast), alpha, beta
if(ng_fit_jast .eq. 1) then
coef_gauss_j_mu_x = (/ -0.47947881d0 /)
expo_gauss_j_mu_x = (/ 3.4987848d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
elseif(ng_fit_jast .eq. 2) then
coef_gauss_j_mu_x = (/ -0.18390742d0, -0.35512656d0 /)
expo_gauss_j_mu_x = (/ 31.9279947d0 , 2.11428789d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
elseif(ng_fit_jast .eq. 3) then
coef_gauss_j_mu_x = (/ -0.07501725d0, -0.28499012d0, -0.1953932d0 /)
expo_gauss_j_mu_x = (/ 206.74058566d0, 1.72974157d0, 11.18735164d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
elseif(ng_fit_jast .eq. 5) then
coef_gauss_j_mu_x = (/ -0.01832955d0 , -0.10188952d0 , -0.20710858d0 , -0.18975032d0 , -0.04641657d0 /)
expo_gauss_j_mu_x = (/ 4.33116687d+03, 2.61292842d+01, 1.43447161d+00, 4.92767426d+00, 2.10654699d+02 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
elseif(ng_fit_jast .eq. 6) then
coef_gauss_j_mu_x = (/ -0.08783664d0 , -0.16088711d0 , -0.18464486d0 , -0.0368509d0 , -0.08130028d0 , -0.0126972d0 /)
expo_gauss_j_mu_x = (/ 4.09729729d+01, 7.11620618d+00, 2.03692338d+00, 4.10831731d+02, 1.12480198d+00, 1.00000000d+04 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
elseif(ng_fit_jast .eq. 7) then
coef_gauss_j_mu_x = (/ -0.01756495d0 , -0.01023623d0 , -0.06548959d0 , -0.03539446d0 , -0.17150646d0 , -0.15071096d0 , -0.11326834d0 /)
expo_gauss_j_mu_x = (/ 9.88572565d+02, 1.21363371d+04, 3.69794870d+01, 1.67364529d+02, 3.03962934d+00, 1.27854005d+00, 9.76383343d+00 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
elseif(ng_fit_jast .eq. 8) then
coef_gauss_j_mu_x = (/ -0.11489205d0 , -0.16008968d0 , -0.12892456d0 , -0.04250838d0 , -0.0718451d0 , -0.02394051d0 , -0.00913353d0 , -0.01285182d0 /)
expo_gauss_j_mu_x = (/ 6.97632442d+00, 2.56010878d+00, 1.22760977d+00, 7.47697124d+01, 2.16104215d+01, 2.96549728d+02, 1.40773328d+04, 1.43335159d+03 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
enddo
!elseif(ng_fit_jast .eq. 9) then
! coef_gauss_j_mu_x = (/ /)
! expo_gauss_j_mu_x = (/ /)
! tmp = mu_erf * mu_erf
! do i = 1, ng_fit_jast
! expo_gauss_j_mu_x(i) = tmp * expo_gauss_j_mu_x(i)
! enddo
elseif(ng_fit_jast .eq. 20) then
ASSERT(n_max_fit_slat == 20)
alpha = expo_j_xmu(1) * mu_erf
call expo_fit_slater_gam(alpha, expos)
beta = expo_j_xmu(2) * mu_erf * mu_erf
tmp = -1.0d0 / sqrt(dacos(-1.d0))
do i = 1, ng_fit_jast
expo_gauss_j_mu_x(i) = expos(i) + beta
coef_gauss_j_mu_x(i) = tmp * coef_fit_slat_gauss(i)
enddo
else
print *, ' not implemented yet'
stop
endif
tmp = 0.5d0 / mu_erf
do i = 1, ng_fit_jast
coef_gauss_j_mu_x(i) = tmp * coef_gauss_j_mu_x(i)
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, expo_gauss_j_mu_x_2, (ng_fit_jast)]
&BEGIN_PROVIDER [double precision, coef_gauss_j_mu_x_2, (ng_fit_jast)]
BEGIN_DOC
!
! J(mu,r12)^2 = 0.25/mu^2 F(r12*mu)^2
!
! F(x)^2 = 1/pi * exp(-2 * alpha * x) exp(-2 * beta * x^2)
!
! The slater function exp(-2 * alpha * x) is fitted with n_max_fit_slat gaussians
!
! See Appendix 2 of JCP 154, 084119 (2021)
!
END_DOC
implicit none
integer :: i
double precision :: tmp
double precision :: expos(ng_fit_jast), alpha, beta
double precision :: alpha_opt, beta_opt
if(ng_fit_jast .eq. 1) then
coef_gauss_j_mu_x_2 = (/ 0.26699573d0 /)
expo_gauss_j_mu_x_2 = (/ 11.71029824d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
elseif(ng_fit_jast .eq. 2) then
coef_gauss_j_mu_x_2 = (/ 0.11627934d0 , 0.18708824d0 /)
expo_gauss_j_mu_x_2 = (/ 102.41386863d0, 6.36239771d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
elseif(ng_fit_jast .eq. 3) then
coef_gauss_j_mu_x_2 = (/ 0.04947216d0 , 0.14116238d0, 0.12276501d0 /)
expo_gauss_j_mu_x_2 = (/ 635.29701766d0, 4.87696954d0, 33.36745891d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
elseif(ng_fit_jast .eq. 5) then
coef_gauss_j_mu_x_2 = (/ 0.01461527d0 , 0.03257147d0 , 0.08831354d0 , 0.11411794d0 , 0.06858783d0 /)
expo_gauss_j_mu_x_2 = (/ 8.76554470d+03, 4.90224577d+02, 3.68267125d+00, 1.29663940d+01, 6.58240931d+01 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
elseif(ng_fit_jast .eq. 6) then
coef_gauss_j_mu_x_2 = (/ 0.01347632d0 , 0.03929124d0 , 0.06289468d0 , 0.10702493d0 , 0.06999865d0 , 0.02558191d0 /)
expo_gauss_j_mu_x_2 = (/ 1.00000000d+04, 1.20900717d+02, 3.20346191d+00, 8.92157196d+00, 3.28119120d+01, 6.49045808d+02 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
elseif(ng_fit_jast .eq. 7) then
coef_gauss_j_mu_x_2 = (/ 0.05202849d0 , 0.01031081d0 , 0.04699157d0 , 0.01451002d0 , 0.07442576d0 , 0.02692033d0 , 0.09311842d0 /)
expo_gauss_j_mu_x_2 = (/ 3.04469415d+00, 1.40682034d+04, 7.45960945d+01, 1.43067466d+03, 2.16815661d+01, 2.95750306d+02, 7.23471236d+00 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
elseif(ng_fit_jast .eq. 8) then
coef_gauss_j_mu_x_2 = (/ 0.00942115d0 , 0.07332421d0 , 0.0508308d0 , 0.08204949d0 , 0.0404099d0 , 0.03201288d0 , 0.01911313d0 , 0.01114732d0 /)
expo_gauss_j_mu_x_2 = (/ 1.56957321d+04, 1.52867810d+01, 4.36016903d+01, 5.96818956d+00, 2.85535269d+00, 1.36064008d+02, 4.71968910d+02, 1.92022350d+03 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
enddo
!elseif(ng_fit_jast .eq. 9) then
! coef_gauss_j_mu_x_2 = (/ /)
! expo_gauss_j_mu_x_2 = (/ /)
!
! tmp = mu_erf * mu_erf
! do i = 1, ng_fit_jast
! expo_gauss_j_mu_x_2(i) = tmp * expo_gauss_j_mu_x_2(i)
! enddo
elseif(ng_fit_jast .eq. 20) then
ASSERT(n_max_fit_slat == 20)
!alpha_opt = 2.d0 * expo_j_xmu(1)
!beta_opt = 2.d0 * expo_j_xmu(2)
! direct opt
alpha_opt = 3.52751759d0
beta_opt = 1.26214809d0
alpha = alpha_opt * mu_erf
call expo_fit_slater_gam(alpha, expos)
beta = beta_opt * mu_erf * mu_erf
tmp = 1.d0 / dacos(-1.d0)
do i = 1, ng_fit_jast
expo_gauss_j_mu_x_2(i) = expos(i) + beta
coef_gauss_j_mu_x_2(i) = tmp * coef_fit_slat_gauss(i)
enddo
else
print *, ' not implemented yet'
stop
endif
tmp = 0.25d0 / (mu_erf * mu_erf)
do i = 1, ng_fit_jast
coef_gauss_j_mu_x_2(i) = tmp * coef_gauss_j_mu_x_2(i)
enddo
END_PROVIDER
! ---
BEGIN_PROVIDER [double precision, expo_gauss_j_mu_1_erf, (ng_fit_jast)]
&BEGIN_PROVIDER [double precision, coef_gauss_j_mu_1_erf, (ng_fit_jast)]
BEGIN_DOC
!
! J(mu,r12) x \frac{1 - erf(mu * r12)}{2} =
!
! - \frac{1}{4 \sqrt{\pi} \mu} \exp(-(alpha1 + alpha2) * mu * r12 - (beta1 + beta2) * mu^2 * r12^2)
!
END_DOC
implicit none
integer :: i
double precision :: tmp
double precision :: expos(ng_fit_jast), alpha, beta
double precision :: alpha_opt, beta_opt
if(ng_fit_jast .eq. 1) then
coef_gauss_j_mu_1_erf = (/ -0.47742461d0 /)
expo_gauss_j_mu_1_erf = (/ 8.72255696d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
elseif(ng_fit_jast .eq. 2) then
coef_gauss_j_mu_1_erf = (/ -0.19342649d0, -0.34563835d0 /)
expo_gauss_j_mu_1_erf = (/ 78.66099999d0, 5.04324363d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
elseif(ng_fit_jast .eq. 3) then
coef_gauss_j_mu_1_erf = (/ -0.0802541d0 , -0.27019258d0, -0.20546681d0 /)
expo_gauss_j_mu_1_erf = (/ 504.53350764d0, 4.01408169d0, 26.5758329d0 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
elseif(ng_fit_jast .eq. 5) then
coef_gauss_j_mu_1_erf = (/ -0.02330531d0 , -0.11888176d0 , -0.16476192d0 , -0.19874713d0 , -0.05889174d0 /)
expo_gauss_j_mu_1_erf = (/ 1.00000000d+04, 4.66067922d+01, 3.04359857d+00, 9.54726649d+00, 3.59796835d+02 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
elseif(ng_fit_jast .eq. 6) then
coef_gauss_j_mu_1_erf = (/ -0.01865654d0 , -0.18319251d0 , -0.06543196d0 , -0.11522778d0 , -0.14825793d0 , -0.03327101d0 /)
expo_gauss_j_mu_1_erf = (/ 1.00000000d+04, 8.05593848d+00, 1.27986190d+02, 2.92674319d+01, 2.93583623d+00, 7.65609148d+02 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
elseif(ng_fit_jast .eq. 7) then
coef_gauss_j_mu_1_erf = (/ -0.11853067d0 , -0.01522824d0 , -0.07419098d0 , -0.022202d0 , -0.12242283d0 , -0.04177571d0 , -0.16983107d0 /)
expo_gauss_j_mu_1_erf = (/ 2.74057056d+00, 1.37626591d+04, 6.65578663d+01, 1.34693031d+03, 1.90547699d+01, 2.69445390d+02, 6.31845879d+00/)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
elseif(ng_fit_jast .eq. 8) then
coef_gauss_j_mu_1_erf = (/ -0.12263328d0 , -0.04965255d0 , -0.15463564d0 , -0.09675781d0 , -0.0807023d0 , -0.02923298d0 , -0.01381381d0 , -0.01675923d0 /)
expo_gauss_j_mu_1_erf = (/ 1.36101994d+01, 1.24908367d+02, 5.29061388d+00, 2.60692516d+00, 3.93396935d+01, 4.43071610d+02, 1.54902240d+04, 1.85170446d+03 /)
tmp = mu_erf * mu_erf
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
enddo
!elseif(ng_fit_jast .eq. 9) then
! coef_gauss_j_mu_1_erf = (/ /)
! expo_gauss_j_mu_1_erf = (/ /)
! tmp = mu_erf * mu_erf
! do i = 1, ng_fit_jast
! expo_gauss_j_mu_1_erf(i) = tmp * expo_gauss_j_mu_1_erf(i)
! enddo
elseif(ng_fit_jast .eq. 20) then
ASSERT(n_max_fit_slat == 20)
!alpha_opt = expo_j_xmu(1) + expo_gauss_1_erf_x(1)
!beta_opt = expo_j_xmu(2) + expo_gauss_1_erf_x(2)
! direct opt
alpha_opt = 2.87875632d0
beta_opt = 1.34801003d0
alpha = alpha_opt * mu_erf
call expo_fit_slater_gam(alpha, expos)
beta = beta_opt * mu_erf * mu_erf
tmp = -1.d0 / dsqrt(dacos(-1.d0))
do i = 1, ng_fit_jast
expo_gauss_j_mu_1_erf(i) = expos(i) + beta
coef_gauss_j_mu_1_erf(i) = tmp * coef_fit_slat_gauss(i)
enddo
else
print *, ' not implemented yet'
stop
endif
tmp = 0.25d0 / mu_erf
do i = 1, ng_fit_jast
coef_gauss_j_mu_1_erf(i) = tmp * coef_gauss_j_mu_1_erf(i)
enddo
END_PROVIDER
! ---
double precision function F_x_j(x)
implicit none
BEGIN_DOC
! F_x_j(x) = dimension-less correlation factor = x (1 - erf(x)) - 1/sqrt(pi) exp(-x^2)
END_DOC
double precision, intent(in) :: x
F_x_j = x * (1.d0 - derf(x)) - 1/dsqrt(dacos(-1.d0)) * dexp(-x**2)
end
double precision function j_mu_F_x_j(x)
implicit none
BEGIN_DOC
! j_mu_F_x_j(x) = correlation factor = 1/2 r12 * (1 - erf(mu*r12)) - 1/(2 sqrt(pi)*mu) exp(-(mu*r12)^2)
!
! = 1/(2*mu) * F_x_j(mu*x)
END_DOC
double precision :: F_x_j
double precision, intent(in) :: x
j_mu_F_x_j = 0.5d0/mu_erf * F_x_j(x*mu_erf)
end
double precision function j_mu(x)
implicit none
double precision, intent(in) :: x
BEGIN_DOC
! j_mu(x) = correlation factor = 1/2 r12 * (1 - erf(mu*r12)) - 1/(2 sqrt(pi)*mu) exp(-(mu*r12)^2)
END_DOC
j_mu = 0.5d0* x * (1.d0 - derf(mu_erf*x)) - 0.5d0/( dsqrt(dacos(-1.d0))*mu_erf) * dexp(-(mu_erf*x)*(mu_erf*x))
end
double precision function j_mu_fit_gauss(x)
implicit none
BEGIN_DOC
! j_mu_fit_gauss(x) = correlation factor = 1/2 r12 * (1 - erf(mu*r12)) - 1/(2 sqrt(pi)*mu) exp(-(mu*r12)^2)
!
! but fitted with gaussians
END_DOC
double precision, intent(in) :: x
integer :: i
double precision :: alpha,coef
j_mu_fit_gauss = 0.d0
do i = 1, n_max_fit_slat
alpha = expo_gauss_j_mu_x(i)
coef = coef_gauss_j_mu_x(i)
j_mu_fit_gauss += coef * dexp(-alpha*x*x)
enddo
end
! ---