# Copyright 2014-2022 The PySCF Developers. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# Author: Abdelrahman Ahmed <>
# Samragni Banerjee <samragnibanerjee4@gmail.com>
# James Serna <jamcar456@gmail.com>
# Terrence Stahl <>
# Alexander Sokolov <alexander.y.sokolov@gmail.com>
#
'''
Unrestricted algebraic diagrammatic construction
'''
import numpy as np
from pyscf import lib, symm
from pyscf.lib import logger
from pyscf.adc import uadc
from pyscf.adc import uadc_ao2mo
from pyscf.adc import radc_ao2mo
from pyscf.adc import dfadc
from pyscf import __config__
from pyscf import df
[docs]
def get_imds(adc, eris=None):
cput0 = (logger.process_clock(), logger.perf_counter())
log = logger.Logger(adc.stdout, adc.verbose)
if adc.method not in ("adc(2)", "adc(2)-x", "adc(3)"):
raise NotImplementedError(adc.method)
method = adc.method
t1 = adc.t1
t2 = adc.t2
nocc_a = adc.nocc_a
nocc_b = adc.nocc_b
e_occ_a = adc.mo_energy_a[:nocc_a]
e_occ_b = adc.mo_energy_b[:nocc_b]
idn_occ_a = np.identity(nocc_a)
idn_occ_b = np.identity(nocc_b)
if eris is None:
eris = adc.transform_integrals()
eris_ovvo = eris.ovvo
eris_OVVO = eris.OVVO
eris_ovVO = eris.ovVO
eris_OVvo = eris.OVvo
# i-j block
# Zeroth-order terms
M_ij_a = lib.einsum('ij,j->ij', idn_occ_a ,e_occ_a)
M_ij_b = lib.einsum('ij,j->ij', idn_occ_b ,e_occ_b)
# Second-order terms
t2_1_a = t2[0][0][:]
M_ij_a += 0.5 * 0.5 * lib.einsum('ilde,jdel->ij',t2_1_a, eris_ovvo, optimize=True)
M_ij_a -= 0.5 * 0.5 * lib.einsum('ilde,jedl->ij',t2_1_a, eris_ovvo, optimize=True)
M_ij_a += 0.5 * 0.5 * lib.einsum('jlde,idel->ij',t2_1_a, eris_ovvo, optimize=True)
M_ij_a -= 0.5 * 0.5 * lib.einsum('jlde,ldei->ij',t2_1_a, eris_ovvo, optimize=True)
t2_1_b = t2[0][2][:]
M_ij_b += 0.5 * 0.5 * lib.einsum('ilde,jdel->ij',t2_1_b, eris_OVVO, optimize=True)
M_ij_b -= 0.5 * 0.5 * lib.einsum('ilde,jedl->ij',t2_1_b, eris_OVVO, optimize=True)
M_ij_b += 0.5 * 0.5 * lib.einsum('jlde,idel->ij',t2_1_b, eris_OVVO, optimize=True)
M_ij_b -= 0.5 * 0.5 * lib.einsum('jlde,ldei->ij',t2_1_b, eris_OVVO, optimize=True)
t2_1_ab = t2[0][1][:]
M_ij_a += 0.5 * lib.einsum('ilde,jdel->ij',t2_1_ab, eris_ovVO, optimize=True)
M_ij_b += 0.5 * lib.einsum('lied,ledj->ij',t2_1_ab, eris_ovVO, optimize=True)
M_ij_a += 0.5 * lib.einsum('jlde,idel->ij',t2_1_ab, eris_ovVO, optimize=True)
M_ij_b += 0.5 * lib.einsum('ljed,ledi->ij',t2_1_ab, eris_ovVO, optimize=True)
del t2_1_ab
# Third-order terms
if (method == "adc(3)"):
t1_2_a, t1_2_b = t1[0]
eris_oovv = eris.oovv
eris_OOVV = eris.OOVV
eris_ooVV = eris.ooVV
eris_OOvv = eris.OOvv
eris_ovvo = eris.ovvo
eris_OVVO = eris.OVVO
eris_ovVO = eris.ovVO
eris_OVvo = eris.OVvo
eris_ovoo = eris.ovoo
eris_OVOO = eris.OVOO
eris_ovOO = eris.ovOO
eris_OVoo = eris.OVoo
eris_oooo = eris.oooo
eris_OOOO = eris.OOOO
eris_ooOO = eris.ooOO
M_ij_a += lib.einsum('ld,ldji->ij',t1_2_a, eris_ovoo, optimize=True)
M_ij_a -= lib.einsum('ld,jdli->ij',t1_2_a, eris_ovoo, optimize=True)
M_ij_a += lib.einsum('ld,ldji->ij',t1_2_b, eris_OVoo, optimize=True)
M_ij_b += lib.einsum('ld,ldji->ij',t1_2_b, eris_OVOO, optimize=True)
M_ij_b -= lib.einsum('ld,jdli->ij',t1_2_b, eris_OVOO, optimize=True)
M_ij_b += lib.einsum('ld,ldji->ij',t1_2_a, eris_ovOO, optimize=True)
M_ij_a += lib.einsum('ld,ldij->ij',t1_2_a, eris_ovoo, optimize=True)
M_ij_a -= lib.einsum('ld,idlj->ij',t1_2_a, eris_ovoo, optimize=True)
M_ij_a += lib.einsum('ld,ldij->ij',t1_2_b, eris_OVoo, optimize=True)
M_ij_b += lib.einsum('ld,ldij->ij',t1_2_b, eris_OVOO, optimize=True)
M_ij_b -= lib.einsum('ld,idlj->ij',t1_2_b, eris_OVOO, optimize=True)
M_ij_b += lib.einsum('ld,ldij->ij',t1_2_a, eris_ovOO, optimize=True)
t2_1_a = t2[0][0][:]
t2_2_a = t2[1][0][:]
M_ij_a += 0.5 * 0.5* lib.einsum('ilde,jdel->ij',t2_2_a, eris_ovvo, optimize=True)
M_ij_a -= 0.5 * 0.5* lib.einsum('ilde,jedl->ij',t2_2_a, eris_ovvo, optimize=True)
M_ij_a += 0.5 * 0.5* lib.einsum('jlde,ledi->ij',t2_2_a, eris_ovvo, optimize=True)
M_ij_a -= 0.5 * 0.5* lib.einsum('jlde,iedl->ij',t2_2_a, eris_ovvo, optimize=True)
t2_1_b = t2[0][2][:]
t2_2_b = t2[1][2][:]
M_ij_b += 0.5 * 0.5* lib.einsum('ilde,jdel->ij',t2_2_b, eris_OVVO, optimize=True)
M_ij_b -= 0.5 * 0.5* lib.einsum('ilde,jedl->ij',t2_2_b, eris_OVVO, optimize=True)
M_ij_b += 0.5 * 0.5* lib.einsum('jlde,ledi->ij',t2_2_b, eris_OVVO, optimize=True)
M_ij_b -= 0.5 * 0.5* lib.einsum('jlde,iedl->ij',t2_2_b, eris_OVVO, optimize=True)
t2_1_ab = t2[0][1][:]
t2_2_ab = t2[1][1][:]
M_ij_a += 0.5 * lib.einsum('ilde,jdel->ij',t2_2_ab, eris_ovVO, optimize=True)
M_ij_b += 0.5 * lib.einsum('lied,ledj->ij',t2_2_ab, eris_ovVO, optimize=True)
M_ij_a += 0.5 * lib.einsum('jlde,ledi->ij',t2_2_ab, eris_OVvo, optimize=True)
M_ij_b += 0.5 * lib.einsum('ljed,ledi->ij',t2_2_ab, eris_ovVO, optimize=True)
t2_1_a = t2[0][0][:]
M_ij_t = lib.einsum('lmde,jldf->mejf', t2_1_a, t2_1_a, optimize=True)
M_ij_a -= 0.5 * lib.einsum('mejf,mefi->ij',M_ij_t, eris_ovvo, optimize=True)
M_ij_a -= 0.5 * lib.einsum('mejf,mefi->ji',M_ij_t, eris_ovvo, optimize=True)
M_ij_a += 0.5 * lib.einsum('mejf,mife->ij',M_ij_t, eris_oovv, optimize=True)
M_ij_a += 0.5 * lib.einsum('mejf,mife->ji',M_ij_t, eris_oovv, optimize=True)
del M_ij_t
M_ij_a += 0.5 * 0.25*lib.einsum('lmde,jnde,limn->ij',t2_1_a,
t2_1_a,eris_oooo, optimize=True)
M_ij_a -= 0.5 * 0.25*lib.einsum('lmde,jnde,lnmi->ij',t2_1_a,
t2_1_a,eris_oooo, optimize=True)
M_ij_a += 0.5 *0.25*lib.einsum('inde,lmde,jlnm->ij',t2_1_a, t2_1_a,eris_oooo, optimize=True)
M_ij_a -= 0.5 *0.25*lib.einsum('inde,lmde,jmnl->ij',t2_1_a, t2_1_a,eris_oooo, optimize=True)
M_ij_a += 0.5*lib.einsum('lmdf,lmde,jief->ij',t2_1_a, t2_1_a, eris_oovv, optimize=True)
M_ij_a -= 0.5*lib.einsum('lmdf,lmde,jfei->ij',t2_1_a, t2_1_a, eris_ovvo, optimize=True)
M_ij_b +=0.5*lib.einsum('lmdf,lmde,jief->ij',t2_1_a, t2_1_a, eris_OOvv , optimize=True)
M_ij_a -= 0.5*lib.einsum('lnde,lmde,jinm->ij',t2_1_a, t2_1_a, eris_oooo, optimize=True)
M_ij_a += 0.5*lib.einsum('lnde,lmde,jmni->ij',t2_1_a, t2_1_a, eris_oooo, optimize=True)
M_ij_b -= 0.5 * lib.einsum('lnde,lmde,nmji->ij',t2_1_a, t2_1_a, eris_ooOO, optimize=True)
t2_1_ab = t2[0][1][:]
M_ij_a -= 0.5 * lib.einsum('lmde,jldf,mefi->ij',t2_1_ab, t2_1_a, eris_OVvo,optimize=True)
M_ij_b += 0.5 * lib.einsum('lmde,ljdf,mefi->ij',t2_1_a, t2_1_ab, eris_ovVO ,optimize=True)
M_ij_a -= 0.5 * lib.einsum('lmde,ildf,mefj->ij',t2_1_ab, t2_1_a, eris_OVvo,optimize=True)
M_ij_b += 0.5 * lib.einsum('lmde,lidf,mefj->ij',t2_1_a, t2_1_ab, eris_ovVO ,optimize=True)
del t2_1_a
t2_1_b = t2[0][2][:]
M_ij_a += 0.5 * lib.einsum('lmde,jlfd,mefi->ij',t2_1_b, t2_1_ab, eris_OVvo ,optimize=True)
M_ij_b -= 0.5 * lib.einsum('mled,jldf,mefi->ij',t2_1_ab, t2_1_b, eris_ovVO,optimize=True)
M_ij_a += 0.5 * lib.einsum('lmde,ilfd,mefj->ij',t2_1_b, t2_1_ab, eris_OVvo ,optimize=True)
M_ij_b -= 0.5 * lib.einsum('mled,ildf,mefj->ij',t2_1_ab, t2_1_b, eris_ovVO,optimize=True)
M_ij_t = lib.einsum('lmde,jldf->mejf', t2_1_b, t2_1_b, optimize=True)
M_ij_b -= 0.5 * lib.einsum('mejf,mefi->ij',M_ij_t, eris_OVVO, optimize=True)
M_ij_b -= 0.5 * lib.einsum('mejf,mefi->ji',M_ij_t, eris_OVVO, optimize=True)
M_ij_b += 0.5 * lib.einsum('mejf,mife->ij',M_ij_t, eris_OOVV, optimize=True)
M_ij_b += 0.5 * lib.einsum('mejf,mife->ji',M_ij_t, eris_OOVV, optimize=True)
del M_ij_t
M_ij_b += 0.5 * 0.25*lib.einsum('lmde,jnde,limn->ij',t2_1_b,
t2_1_b,eris_OOOO, optimize=True)
M_ij_b -= 0.5 * 0.25*lib.einsum('lmde,jnde,lnmi->ij',t2_1_b,
t2_1_b,eris_OOOO, optimize=True)
M_ij_b += 0.5 * 0.25*lib.einsum('inde,lmde,jlnm->ij',t2_1_b,
t2_1_b,eris_OOOO, optimize=True)
M_ij_b -= 0.5 * 0.25*lib.einsum('inde,lmde,jmnl->ij',t2_1_b,
t2_1_b,eris_OOOO, optimize=True)
M_ij_a += 0.5*lib.einsum('lmdf,lmde,jief->ij',t2_1_b, t2_1_b, eris_ooVV , optimize=True)
M_ij_b += 0.5*lib.einsum('lmdf,lmde,jief->ij',t2_1_b, t2_1_b, eris_OOVV , optimize=True)
M_ij_b -= 0.5*lib.einsum('lmdf,lmde,jfei->ij',t2_1_b, t2_1_b, eris_OVVO , optimize=True)
M_ij_a -= 0.5 * lib.einsum('lnde,lmde,jinm->ij',t2_1_b, t2_1_b, eris_ooOO, optimize=True)
M_ij_b -= 0.5*lib.einsum('lnde,lmde,jinm->ij',t2_1_b, t2_1_b, eris_OOOO, optimize=True)
M_ij_b += 0.5*lib.einsum('lnde,lmde,jmni->ij',t2_1_b, t2_1_b, eris_OOOO, optimize=True)
del t2_1_b
t2_1_ab = t2[0][1][:]
M_ij_a += 0.5 * lib.einsum('mled,jlfd,mefi->ij',t2_1_ab, t2_1_ab, eris_ovvo ,optimize=True)
M_ij_a -= 0.5 * lib.einsum('mled,jlfd,mife->ij',t2_1_ab, t2_1_ab, eris_oovv ,optimize=True)
M_ij_a -= 0.5 * lib.einsum('mlde,jldf,mife->ij',t2_1_ab, t2_1_ab, eris_ooVV ,optimize=True)
M_ij_b += 0.5 * lib.einsum('lmde,ljdf,mefi->ij',t2_1_ab, t2_1_ab, eris_OVVO,optimize=True)
M_ij_b -= 0.5 * lib.einsum('lmde,ljdf,mife->ij',t2_1_ab, t2_1_ab, eris_OOVV,optimize=True)
M_ij_b -= 0.5 * lib.einsum('lmed,ljfd,mife->ij',t2_1_ab, t2_1_ab, eris_OOvv ,optimize=True)
M_ij_a += 0.5 * lib.einsum('mled,ilfd,mefj->ij',t2_1_ab, t2_1_ab, eris_ovvo ,optimize=True)
M_ij_a -= 0.5 * lib.einsum('mled,ilfd,mjfe->ij',t2_1_ab, t2_1_ab, eris_oovv ,optimize=True)
M_ij_a -= 0.5 * lib.einsum('mlde,ildf,mjfe->ij',t2_1_ab, t2_1_ab, eris_ooVV ,optimize=True)
M_ij_b += 0.5 * lib.einsum('lmde,lidf,mefj->ij',t2_1_ab, t2_1_ab, eris_OVVO ,optimize=True)
M_ij_b -= 0.5 * lib.einsum('lmde,lidf,mjfe->ij',t2_1_ab, t2_1_ab, eris_OOVV ,optimize=True)
M_ij_b -= 0.5 * lib.einsum('lmed,lifd,mjfe->ij',t2_1_ab, t2_1_ab, eris_OOvv ,optimize=True)
M_ij_a += 0.5 * lib.einsum('lmde,jnde,limn->ij',t2_1_ab ,t2_1_ab,eris_ooOO, optimize=True)
M_ij_b += 0.5 * lib.einsum('mled,njed,mnli->ij',t2_1_ab ,t2_1_ab,eris_ooOO, optimize=True)
M_ij_a +=0.5 * lib.einsum('inde,lmde,jlnm->ij',t2_1_ab, t2_1_ab,eris_ooOO, optimize=True)
M_ij_b +=0.5 * lib.einsum('nied,mled,nmjl->ij',t2_1_ab, t2_1_ab,eris_ooOO, optimize=True)
M_ij_a +=lib.einsum('mlfd,mled,jief->ij',t2_1_ab, t2_1_ab, eris_oovv , optimize=True)
M_ij_a -=lib.einsum('mlfd,mled,jfei->ij',t2_1_ab, t2_1_ab, eris_ovvo , optimize=True)
M_ij_a +=lib.einsum('lmdf,lmde,jief->ij',t2_1_ab, t2_1_ab, eris_ooVV , optimize=True)
M_ij_b +=lib.einsum('lmdf,lmde,jief->ij',t2_1_ab, t2_1_ab, eris_OOVV , optimize=True)
M_ij_b -=lib.einsum('lmdf,lmde,jfei->ij',t2_1_ab, t2_1_ab, eris_OVVO , optimize=True)
M_ij_b +=lib.einsum('lmfd,lmed,jief->ij',t2_1_ab, t2_1_ab, eris_OOvv , optimize=True)
M_ij_a -= lib.einsum('nled,mled,jinm->ij',t2_1_ab, t2_1_ab, eris_oooo, optimize=True)
M_ij_a += lib.einsum('nled,mled,jmni->ij',t2_1_ab, t2_1_ab, eris_oooo, optimize=True)
M_ij_a -= lib.einsum('lnde,lmde,jinm->ij',t2_1_ab, t2_1_ab, eris_ooOO, optimize=True)
M_ij_b -= lib.einsum('lnde,lmde,jinm->ij',t2_1_ab, t2_1_ab, eris_OOOO, optimize=True)
M_ij_b += lib.einsum('lnde,lmde,jmni->ij',t2_1_ab, t2_1_ab, eris_OOOO, optimize=True)
M_ij_b -= lib.einsum('nled,mled,nmji->ij',t2_1_ab, t2_1_ab, eris_ooOO, optimize=True)
del t2_1_ab
M_ij = (M_ij_a, M_ij_b)
cput0 = log.timer_debug1("Completed M_ab ADC(3) calculation", *cput0)
return M_ij
[docs]
def get_diag(adc,M_ij=None,eris=None):
if adc.method not in ("adc(2)", "adc(2)-x", "adc(3)"):
raise NotImplementedError(adc.method)
if M_ij is None:
M_ij = adc.get_imds()
M_ij_a, M_ij_b = M_ij[0], M_ij[1]
nocc_a = adc.nocc_a
nocc_b = adc.nocc_b
nvir_a = adc.nvir_a
nvir_b = adc.nvir_b
n_singles_a = nocc_a
n_singles_b = nocc_b
n_doubles_aaa = nocc_a * (nocc_a - 1) * nvir_a // 2
n_doubles_bab = nvir_b * nocc_a * nocc_b
n_doubles_aba = nvir_a * nocc_b * nocc_a
n_doubles_bbb = nocc_b * (nocc_b - 1) * nvir_b // 2
dim = n_singles_a + n_singles_b + n_doubles_aaa + n_doubles_bab + n_doubles_aba + n_doubles_bbb
e_occ_a = adc.mo_energy_a[:nocc_a]
e_occ_b = adc.mo_energy_b[:nocc_b]
e_vir_a = adc.mo_energy_a[nocc_a:]
e_vir_b = adc.mo_energy_b[nocc_b:]
ij_ind_a = np.tril_indices(nocc_a, k=-1)
ij_ind_b = np.tril_indices(nocc_b, k=-1)
s_a = 0
f_a = n_singles_a
s_b = f_a
f_b = s_b + n_singles_b
s_aaa = f_b
f_aaa = s_aaa + n_doubles_aaa
s_bab = f_aaa
f_bab = s_bab + n_doubles_bab
s_aba = f_bab
f_aba = s_aba + n_doubles_aba
s_bbb = f_aba
f_bbb = s_bbb + n_doubles_bbb
d_ij_a = e_occ_a[:,None] + e_occ_a
d_a_a = e_vir_a[:,None]
D_n_a = -d_a_a + d_ij_a.reshape(-1)
D_n_a = D_n_a.reshape((nvir_a,nocc_a,nocc_a))
D_aij_a = D_n_a.copy()[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
d_ij_b = e_occ_b[:,None] + e_occ_b
d_a_b = e_vir_b[:,None]
D_n_b = -d_a_b + d_ij_b.reshape(-1)
D_n_b = D_n_b.reshape((nvir_b,nocc_b,nocc_b))
D_aij_b = D_n_b.copy()[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
d_ij_ab = e_occ_b[:,None] + e_occ_a
d_a_b = e_vir_b[:,None]
D_n_bab = -d_a_b + d_ij_ab.reshape(-1)
D_aij_bab = D_n_bab.reshape(-1)
d_ij_ab = e_occ_a[:,None] + e_occ_b
d_a_a = e_vir_a[:,None]
D_n_aba = -d_a_a + d_ij_ab.reshape(-1)
D_aij_aba = D_n_aba.reshape(-1)
diag = np.zeros(dim)
# Compute precond in h1-h1 block
M_ij_a_diag = np.diagonal(M_ij_a)
M_ij_b_diag = np.diagonal(M_ij_b)
diag[s_a:f_a] = M_ij_a_diag.copy()
diag[s_b:f_b] = M_ij_b_diag.copy()
# Compute precond in 2p1h-2p1h block
diag[s_aaa:f_aaa] = D_aij_a.copy()
diag[s_bab:f_bab] = D_aij_bab.copy()
diag[s_aba:f_aba] = D_aij_aba.copy()
diag[s_bbb:f_bbb] = D_aij_b.copy()
# ###### Additional terms for the preconditioner ####
# if (method == "adc(2)-x" or method == "adc(3)"):
#
# if eris is None:
# eris = adc.transform_integrals()
#
# if isinstance(eris.vvvv_p, np.ndarray):
#
# eris_oooo = eris.oooo
# eris_OOOO = eris.OOOO
# eris_ooOO = eris.ooOO
# eris_oovv = eris.oovv
# eris_OOVV = eris.OOVV
# eris_ooVV = eris.ooVV
# eris_OOvv = eris.OOvv
# eris_ovvo = eris.ovvo
# eris_OVVO = eris.OVVO
#
# eris_oooo_p = np.ascontiguousarray(eris_oooo.transpose(0,2,1,3))
# eris_oooo_p -= np.ascontiguousarray(eris_oooo_p.transpose(0,1,3,2))
# eris_oooo_p = eris_oooo_p.reshape(nocc_a*nocc_a, nocc_a*nocc_a)
#
# temp = np.zeros((nvir_a,eris_oooo_p.shape[0]))
# temp[:] += np.diagonal(eris_oooo_p)
# temp = temp.reshape(nvir_a, nocc_a, nocc_a)
# diag[s_aaa:f_aaa] += -temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
#
# eris_OOOO_p = np.ascontiguousarray(eris_OOOO.transpose(0,2,1,3))
# eris_OOOO_p -= np.ascontiguousarray(eris_OOOO_p.transpose(0,1,3,2))
# eris_OOOO_p = eris_OOOO_p.reshape(nocc_b*nocc_b, nocc_b*nocc_b)
#
# temp = np.zeros((nvir_b,eris_OOOO_p.shape[0]))
# temp[:] += np.diagonal(eris_OOOO_p)
# temp = temp.reshape(nvir_b, nocc_b, nocc_b)
# diag[s_bbb:f_bbb] += -temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
#
# eris_oOoO_p = np.ascontiguousarray(eris_ooOO.transpose(0,2,1,3))
# eris_oOoO_p = eris_oOoO_p.reshape(nocc_a*nocc_b, nocc_a*nocc_b)
#
# temp = np.zeros((nvir_b, eris_oOoO_p.shape[0]))
# temp[:] += np.diag(eris_oOoO_p)
# diag[s_bab:f_bab] += -temp.reshape(-1)
#
# temp = np.zeros((nvir_a, eris_oOoO_p.shape[0]))
# temp[:] += np.diag(eris_oOoO_p.T)
# diag[s_aba:f_aba] += -temp.reshape(-1)
#
# eris_ovov_p = np.ascontiguousarray(eris_oovv.transpose(0,2,1,3))
# eris_ovov_p -= np.ascontiguousarray(eris_ovvo.transpose(0,2,3,1))
# eris_ovov_p = eris_ovov_p.reshape(nocc_a*nvir_a, nocc_a*nvir_a)
#
# temp = np.zeros((nocc_a,nocc_a,nvir_a))
# temp[:] += np.diagonal(eris_ovov_p).reshape(nocc_a,nvir_a)
# temp = np.ascontiguousarray(temp.T)
# diag[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
#
# temp = np.ascontiguousarray(temp.transpose(0,2,1))
# diag[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
#
# eris_OVOV_p = np.ascontiguousarray(eris_OOVV.transpose(0,2,1,3))
# eris_OVOV_p -= np.ascontiguousarray(eris_OVVO.transpose(0,2,3,1))
# eris_OVOV_p = eris_OVOV_p.reshape(nocc_b*nvir_b, nocc_b*nvir_b)
#
# temp = np.zeros((nocc_b,nocc_b,nvir_b))
# temp[:] += np.diagonal(eris_OVOV_p).reshape(nocc_b,nvir_b)
# temp = np.ascontiguousarray(temp.T)
# diag[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
#
# temp = np.ascontiguousarray(temp.transpose(0,2,1))
# diag[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
#
# temp = np.zeros((nocc_a, nocc_b, nvir_b))
# temp[:] += np.diagonal(eris_OVOV_p).reshape(nocc_b, nvir_b)
# temp = np.ascontiguousarray(temp.transpose(2,0,1))
# diag[s_bab:f_bab] += temp.reshape(-1)
#
# temp = np.zeros((nocc_b, nocc_a, nvir_a))
# temp[:] += np.diagonal(eris_ovov_p).reshape(nocc_a, nvir_a)
# temp = np.ascontiguousarray(temp.transpose(2,0,1))
# diag[s_aba:f_aba] += temp.reshape(-1)
#
# eris_oVoV_p = np.ascontiguousarray(eris_ooVV.transpose(0,2,1,3))
# eris_oVoV_p = eris_oVoV_p.reshape(nocc_a*nvir_b, nocc_a*nvir_b)
#
# temp = np.zeros((nocc_b, nocc_a, nvir_b))
# temp[:] += np.diagonal(eris_oVoV_p).reshape(nocc_a,nvir_b)
# temp = np.ascontiguousarray(temp.transpose(2,1,0))
# diag[s_bab:f_bab] += temp.reshape(-1)
#
# eris_OvOv_p = np.ascontiguousarray(eris_OOvv.transpose(0,2,1,3))
# eris_OvOv_p = eris_OvOv_p.reshape(nocc_b*nvir_a, nocc_b*nvir_a)
#
# temp = np.zeros((nocc_a, nocc_b, nvir_a))
# temp[:] += np.diagonal(eris_OvOv_p).reshape(nocc_b,nvir_a)
# temp = np.ascontiguousarray(temp.transpose(2,1,0))
# diag[s_aba:f_aba] += temp.reshape(-1)
# else:
# raise Exception("Precond not available for out-of-core and density-fitted algo")
diag = -diag
return diag
[docs]
def matvec(adc, M_ij=None, eris=None):
if adc.method not in ("adc(2)", "adc(2)-x", "adc(3)"):
raise NotImplementedError(adc.method)
method = adc.method
nocc_a = adc.nocc_a
nocc_b = adc.nocc_b
nvir_a = adc.nvir_a
nvir_b = adc.nvir_b
ij_ind_a = np.tril_indices(nocc_a, k=-1)
ij_ind_b = np.tril_indices(nocc_b, k=-1)
n_singles_a = nocc_a
n_singles_b = nocc_b
n_doubles_aaa = nocc_a * (nocc_a - 1) * nvir_a // 2
n_doubles_bab = nvir_b * nocc_a * nocc_b
n_doubles_aba = nvir_a * nocc_b * nocc_a
n_doubles_bbb = nocc_b * (nocc_b - 1) * nvir_b // 2
dim = n_singles_a + n_singles_b + n_doubles_aaa + n_doubles_bab + n_doubles_aba + n_doubles_bbb
e_occ_a = adc.mo_energy_a[:nocc_a]
e_occ_b = adc.mo_energy_b[:nocc_b]
e_vir_a = adc.mo_energy_a[nocc_a:]
e_vir_b = adc.mo_energy_b[nocc_b:]
if eris is None:
eris = adc.transform_integrals()
d_ij_a = e_occ_a[:,None] + e_occ_a
d_a_a = e_vir_a[:,None]
D_n_a = -d_a_a + d_ij_a.reshape(-1)
D_n_a = D_n_a.reshape((nvir_a,nocc_a,nocc_a))
D_aij_a = D_n_a.copy()[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
d_ij_b = e_occ_b[:,None] + e_occ_b
d_a_b = e_vir_b[:,None]
D_n_b = -d_a_b + d_ij_b.reshape(-1)
D_n_b = D_n_b.reshape((nvir_b,nocc_b,nocc_b))
D_aij_b = D_n_b.copy()[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
d_ij_ab = e_occ_b[:,None] + e_occ_a
d_a_b = e_vir_b[:,None]
D_n_bab = -d_a_b + d_ij_ab.reshape(-1)
D_aij_bab = D_n_bab.reshape(-1)
d_ij_ab = e_occ_a[:,None] + e_occ_b
d_a_a = e_vir_a[:,None]
D_n_aba = -d_a_a + d_ij_ab.reshape(-1)
D_aij_aba = D_n_aba.reshape(-1)
s_a = 0
f_a = n_singles_a
s_b = f_a
f_b = s_b + n_singles_b
s_aaa = f_b
f_aaa = s_aaa + n_doubles_aaa
s_bab = f_aaa
f_bab = s_bab + n_doubles_bab
s_aba = f_bab
f_aba = s_aba + n_doubles_aba
s_bbb = f_aba
f_bbb = s_bbb + n_doubles_bbb
if M_ij is None:
M_ij = adc.get_imds()
M_ij_a, M_ij_b = M_ij
#Calculate sigma vector
def sigma_(r):
cput0 = (logger.process_clock(), logger.perf_counter())
log = logger.Logger(adc.stdout, adc.verbose)
s = np.zeros((dim))
r_a = r[s_a:f_a]
r_b = r[s_b:f_b]
r_aaa = r[s_aaa:f_aaa]
r_bab = r[s_bab:f_bab]
r_aba = r[s_aba:f_aba]
r_bbb = r[s_bbb:f_bbb]
r_aaa = r_aaa.reshape(nvir_a,-1)
r_bbb = r_bbb.reshape(nvir_b,-1)
r_aaa_u = None
r_aaa_u = np.zeros((nvir_a,nocc_a,nocc_a))
r_aaa_u[:,ij_ind_a[0],ij_ind_a[1]]= r_aaa.copy()
r_aaa_u[:,ij_ind_a[1],ij_ind_a[0]]= -r_aaa.copy()
r_bbb_u = None
r_bbb_u = np.zeros((nvir_b,nocc_b,nocc_b))
r_bbb_u[:,ij_ind_b[0],ij_ind_b[1]]= r_bbb.copy()
r_bbb_u[:,ij_ind_b[1],ij_ind_b[0]]= -r_bbb.copy()
r_aba = r_aba.reshape(nvir_a,nocc_a,nocc_b)
r_bab = r_bab.reshape(nvir_b,nocc_b,nocc_a)
eris_ovoo = eris.ovoo
eris_OVOO = eris.OVOO
eris_OVoo = eris.OVoo
eris_ovOO = eris.ovOO
############ ADC(2) ij block ############################
s[s_a:f_a] = lib.einsum('ij,j->i',M_ij_a,r_a)
s[s_b:f_b] = lib.einsum('ij,j->i',M_ij_b,r_b)
############# ADC(2) i - kja block #########################
s[s_a:f_a] += 0.5*lib.einsum('jaki,ajk->i', eris_ovoo, r_aaa_u, optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('kaji,ajk->i', eris_ovoo, r_aaa_u, optimize=True)
s[s_a:f_a] += lib.einsum('jaki,ajk->i', eris_OVoo, r_bab, optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('jaki,ajk->i', eris_OVOO, r_bbb_u, optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('kaji,ajk->i', eris_OVOO, r_bbb_u, optimize=True)
s[s_b:f_b] += lib.einsum('jaki,ajk->i', eris_ovOO, r_aba, optimize=True)
############## ADC(2) ajk - i block ############################
temp = lib.einsum('jaki,i->ajk', eris_ovoo, r_a, optimize=True)
temp -= lib.einsum('kaji,i->ajk', eris_ovoo, r_a, optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
s[s_bab:f_bab] += lib.einsum('jaik,i->ajk', eris_OVoo, r_a, optimize=True).reshape(-1)
s[s_aba:f_aba] += lib.einsum('jaki,i->ajk', eris_ovOO, r_b, optimize=True).reshape(-1)
temp = lib.einsum('jaki,i->ajk', eris_OVOO, r_b, optimize=True)
temp -= lib.einsum('kaji,i->ajk', eris_OVOO, r_b, optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
############ ADC(2) ajk - bil block ############################
r_aaa = r_aaa.reshape(-1)
r_bbb = r_bbb.reshape(-1)
s[s_aaa:f_aaa] += D_aij_a * r_aaa
s[s_bab:f_bab] += D_aij_bab * r_bab.reshape(-1)
s[s_aba:f_aba] += D_aij_aba * r_aba.reshape(-1)
s[s_bbb:f_bbb] += D_aij_b * r_bbb
############### ADC(3) ajk - bil block ############################
if (method == "adc(2)-x" or method == "adc(3)"):
eris_oooo = eris.oooo
eris_OOOO = eris.OOOO
eris_ooOO = eris.ooOO
eris_oovv = eris.oovv
eris_OOVV = eris.OOVV
eris_ooVV = eris.ooVV
eris_OOvv = eris.OOvv
eris_ovvo = eris.ovvo
eris_OVVO = eris.OVVO
eris_ovVO = eris.ovVO
eris_OVvo = eris.OVvo
r_aaa = r_aaa.reshape(nvir_a,-1)
r_bab = r_bab.reshape(nvir_b,nocc_b,nocc_a)
r_aba = r_aba.reshape(nvir_a,nocc_a,nocc_b)
r_bbb = r_bbb.reshape(nvir_b,-1)
r_aaa_u = None
r_aaa_u = np.zeros((nvir_a,nocc_a,nocc_a))
r_aaa_u[:,ij_ind_a[0],ij_ind_a[1]]= r_aaa.copy()
r_aaa_u[:,ij_ind_a[1],ij_ind_a[0]]= -r_aaa.copy()
r_bbb_u = None
r_bbb_u = np.zeros((nvir_b,nocc_b,nocc_b))
r_bbb_u[:,ij_ind_b[0],ij_ind_b[1]]= r_bbb.copy()
r_bbb_u[:,ij_ind_b[1],ij_ind_b[0]]= -r_bbb.copy()
temp = 0.5*lib.einsum('jlki,ail->ajk',eris_oooo,r_aaa_u ,optimize=True)
temp -= 0.5*lib.einsum('jikl,ail->ajk',eris_oooo,r_aaa_u ,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
temp = 0.5*lib.einsum('jlki,ail->ajk',eris_OOOO,r_bbb_u,optimize=True)
temp -= 0.5*lib.einsum('jikl,ail->ajk',eris_OOOO,r_bbb_u,optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
s[s_bab:f_bab] -= 0.5*lib.einsum('kijl,ali->ajk',eris_ooOO,
r_bab,optimize=True).reshape(-1)
s[s_bab:f_bab] -= 0.5*lib.einsum('klji,ail->ajk',eris_ooOO,
r_bab,optimize=True).reshape(-1)
s[s_aba:f_aba] -= 0.5*lib.einsum('jlki,ali->ajk',eris_ooOO,
r_aba,optimize=True).reshape(-1)
s[s_aba:f_aba] -= 0.5*lib.einsum('jikl,ail->ajk',eris_ooOO,
r_aba,optimize=True).reshape(-1)
temp = 0.5*lib.einsum('klba,bjl->ajk',eris_oovv,r_aaa_u,optimize=True)
temp -= 0.5*lib.einsum('kabl,bjl->ajk',eris_ovvo,r_aaa_u,optimize=True)
temp += 0.5* lib.einsum('kabl,blj->ajk',eris_ovVO,r_bab,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
s[s_bab:f_bab] += 0.5*lib.einsum('klba,bjl->ajk',eris_ooVV,
r_bab,optimize=True).reshape(-1)
temp_1 = 0.5*lib.einsum('klba,bjl->ajk',eris_OOVV,r_bbb_u,optimize=True)
temp_1 -= 0.5*lib.einsum('kabl,bjl->ajk',eris_OVVO,r_bbb_u,optimize=True)
temp_1 += 0.5*lib.einsum('kabl,blj->ajk',eris_OVvo,r_aba,optimize=True)
s[s_bbb:f_bbb] += temp_1[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
s[s_aba:f_aba] += 0.5*lib.einsum('klba,bjl->ajk',eris_OOvv,
r_aba,optimize=True).reshape(-1)
temp = -0.5*lib.einsum('jlba,bkl->ajk',eris_oovv,r_aaa_u,optimize=True)
temp += 0.5*lib.einsum('jabl,bkl->ajk',eris_ovvo,r_aaa_u,optimize=True)
temp -= 0.5*lib.einsum('jabl,blk->ajk',eris_ovVO,r_bab,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
s[s_bab:f_bab] += 0.5*lib.einsum('jabl,bkl->ajk',
eris_OVvo,r_aaa_u,optimize=True).reshape(-1)
s[s_bab:f_bab] += 0.5*lib.einsum('jlba,blk->ajk',
eris_OOVV,r_bab,optimize=True).reshape(-1)
s[s_bab:f_bab] -= 0.5*lib.einsum('jabl,blk->ajk',
eris_OVVO,r_bab,optimize=True).reshape(-1)
temp = -0.5*lib.einsum('jlba,bkl->ajk',eris_OOVV,r_bbb_u,optimize=True)
temp += 0.5*lib.einsum('jabl,bkl->ajk',eris_OVVO,r_bbb_u,optimize=True)
temp -= 0.5*lib.einsum('jabl,blk->ajk',eris_OVvo,r_aba,optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
s[s_aba:f_aba] += 0.5*lib.einsum('jlba,blk->ajk',eris_oovv,
r_aba,optimize=True).reshape(-1)
s[s_aba:f_aba] -= 0.5*lib.einsum('jabl,blk->ajk',eris_ovvo,
r_aba,optimize=True).reshape(-1)
s[s_aba:f_aba] += 0.5*lib.einsum('jabl,bkl->ajk',eris_ovVO,
r_bbb_u,optimize=True).reshape(-1)
temp = -0.5*lib.einsum('kiba,bij->ajk',eris_oovv,r_aaa_u,optimize=True)
temp += 0.5*lib.einsum('kabi,bij->ajk',eris_ovvo,r_aaa_u,optimize=True)
temp += 0.5*lib.einsum('kabi,bij->ajk',eris_ovVO,r_bab,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
s[s_bab:f_bab] += 0.5*lib.einsum('kiba,bji->ajk',eris_ooVV,
r_bab,optimize=True).reshape(-1)
temp = -0.5*lib.einsum('kiba,bij->ajk',eris_OOVV,r_bbb_u,optimize=True)
temp += 0.5*lib.einsum('kabi,bij->ajk',eris_OVVO,r_bbb_u,optimize=True)
temp += 0.5*lib.einsum('kabi,bij->ajk',eris_OVvo,r_aba,optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
s[s_aba:f_aba] += 0.5*lib.einsum('kiba,bji->ajk',eris_OOvv,
r_aba,optimize=True).reshape(-1)
temp = 0.5*lib.einsum('jiba,bik->ajk',eris_oovv,r_aaa_u,optimize=True)
temp -= 0.5*lib.einsum('jabi,bik->ajk',eris_ovvo,r_aaa_u,optimize=True)
temp -= 0.5*lib.einsum('jabi,bik->ajk',eris_ovVO,r_bab,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1]].reshape(-1)
s[s_bab:f_bab] += 0.5*lib.einsum('jiba,bik->ajk',eris_OOVV,
r_bab,optimize=True).reshape(-1)
s[s_bab:f_bab] -= 0.5*lib.einsum('jabi,bik->ajk',eris_OVVO,
r_bab,optimize=True).reshape(-1)
s[s_bab:f_bab] -= 0.5*lib.einsum('jabi,bik->ajk',eris_OVvo,
r_aaa_u,optimize=True).reshape(-1)
s[s_aba:f_aba] += 0.5*lib.einsum('jiba,bik->ajk',eris_oovv,
r_aba,optimize=True).reshape(-1)
s[s_aba:f_aba] -= 0.5*lib.einsum('jabi,bik->ajk',eris_ovvo,
r_aba,optimize=True).reshape(-1)
s[s_aba:f_aba] -= 0.5*lib.einsum('jabi,bik->ajk',eris_ovVO,
r_bbb_u,optimize=True).reshape(-1)
temp = 0.5*lib.einsum('jiba,bik->ajk',eris_OOVV,r_bbb_u,optimize=True)
temp -= 0.5*lib.einsum('jabi,bik->ajk',eris_OVVO,r_bbb_u,optimize=True)
temp -= 0.5*lib.einsum('jabi,bik->ajk',eris_OVvo,r_aba,optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1]].reshape(-1)
if (method == "adc(3)"):
eris_ovoo = eris.ovoo
eris_OVOO = eris.OVOO
eris_ovOO = eris.ovOO
eris_OVoo = eris.OVoo
################ ADC(3) i - kja and ajk - i block ############################
t2_1_a = adc.t2[0][0][:]
t2_1_ab = adc.t2[0][1][:]
temp_singles = np.zeros((nocc_a))
temp_doubles = np.zeros((nvir_a, nvir_a, nvir_a))
r_aaa = r_aaa.reshape(nvir_a,-1)
t2_1_a_t = t2_1_a[ij_ind_a[0],ij_ind_a[1],:,:]
temp_1 = lib.einsum('pbc,ap->abc',t2_1_a_t,r_aaa, optimize=True)
if isinstance(eris.ovvv, type(None)):
chnk_size = uadc_ao2mo.calculate_chunk_size(adc)
a = 0
for p in range(0,nocc_a,chnk_size):
eris_ovvv = dfadc.get_ovvv_spin_df(
adc, eris.Lov, eris.Lvv, p, chnk_size).reshape(-1,nvir_a,nvir_a,nvir_a)
k = eris_ovvv.shape[0]
temp_singles[a:a+k] += 0.5* \
lib.einsum('abc,icab->i',temp_1, eris_ovvv, optimize=True)
temp_singles[a:a+k] -= 0.5* \
lib.einsum('abc,ibac->i',temp_1, eris_ovvv, optimize=True)
temp_doubles += lib.einsum('i,icab->bca',r_a[a:a+k],eris_ovvv,optimize=True)
temp_doubles -= lib.einsum('i,ibac->bca',r_a[a:a+k],eris_ovvv,optimize=True)
del eris_ovvv
a += k
else :
eris_ovvv = radc_ao2mo.unpack_eri_1(eris.ovvv, nvir_a)
temp_singles += 0.5*lib.einsum('abc,icab->i',temp_1, eris_ovvv, optimize=True)
temp_singles -= 0.5*lib.einsum('abc,ibac->i',temp_1, eris_ovvv, optimize=True)
temp_doubles += lib.einsum('i,icab->bca',r_a,eris_ovvv,optimize=True)
temp_doubles -= lib.einsum('i,ibac->bca',r_a,eris_ovvv,optimize=True)
del eris_ovvv
s[s_a:f_a] += temp_singles
s[s_aaa:f_aaa] += 0.5*lib.einsum('bca,pbc->ap',temp_doubles,
t2_1_a_t,optimize=True).reshape(-1)
del temp_singles
del temp_doubles
r_aaa_u = np.zeros((nvir_a,nocc_a,nocc_a))
r_aaa_u[:,ij_ind_a[0],ij_ind_a[1]]= r_aaa.copy()
r_aaa_u[:,ij_ind_a[1],ij_ind_a[0]]= -r_aaa.copy()
r_aba = r_aba.reshape(nvir_a,nocc_a,nocc_b)
temp = lib.einsum('jlab,ajk->blk',t2_1_a,r_aaa_u,optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blk,lbik->i',temp,eris_ovoo,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blk,iblk->i',temp,eris_ovoo,optimize=True)
temp_1 = lib.einsum('jlab,ajk->blk',t2_1_a,r_aba,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blk,lbik->i',temp_1,eris_ovOO,optimize=True)
temp = -lib.einsum('klab,akj->blj',t2_1_a,r_aaa_u,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blj,lbij->i',temp,eris_ovoo,optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blj,iblj->i',temp,eris_ovoo,optimize=True)
temp_1 = -lib.einsum('klab,akj->blj',t2_1_a,r_aba,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blj,lbij->i',temp_1,eris_ovOO,optimize=True)
temp_1 = lib.einsum('i,lbik->kbl',r_a, eris_ovoo)
temp_1 -= lib.einsum('i,iblk->kbl',r_a, eris_ovoo)
temp = lib.einsum('kbl,jlab->ajk',temp_1,t2_1_a,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1] ].reshape(-1)
temp_2 = lib.einsum('i,lbik->kbl',r_b,eris_ovOO)
temp = lib.einsum('kbl,jlab->ajk',temp_2,t2_1_a,optimize=True)
s[s_aba:f_aba] += temp.reshape(-1)
temp_1 = lib.einsum('i,lbij->jbl',r_a, eris_ovoo)
temp_1 -= lib.einsum('i,iblj->jbl',r_a, eris_ovoo)
temp = lib.einsum('jbl,klab->ajk',temp_1,t2_1_a,optimize=True)
s[s_aaa:f_aaa] -= temp[:,ij_ind_a[0],ij_ind_a[1] ].reshape(-1)
del t2_1_a
t2_1_b = adc.t2[0][2][:]
temp_singles = np.zeros((nocc_b))
temp_doubles = np.zeros((nvir_b, nvir_b, nvir_b))
r_bbb = r_bbb.reshape(nvir_b,-1)
t2_1_b_t = t2_1_b[ij_ind_b[0],ij_ind_b[1],:,:]
temp_1 = lib.einsum('pbc,ap->abc',t2_1_b_t,r_bbb, optimize=True)
if isinstance(eris.OVVV, type(None)):
chnk_size = uadc_ao2mo.calculate_chunk_size(adc)
a = 0
for p in range(0,nocc_b,chnk_size):
eris_OVVV = dfadc.get_ovvv_spin_df(
adc, eris.LOV, eris.LVV, p, chnk_size).reshape(-1,nvir_b,nvir_b,nvir_b)
k = eris_OVVV.shape[0]
temp_singles[a:a+k] += 0.5* \
lib.einsum('abc,icab->i',temp_1, eris_OVVV, optimize=True)
temp_singles[a:a+k] -= 0.5* \
lib.einsum('abc,ibac->i',temp_1, eris_OVVV, optimize=True)
temp_doubles += lib.einsum('i,icab->bca',r_b[a:a+k],eris_OVVV,optimize=True)
temp_doubles -= lib.einsum('i,ibac->bca',r_b[a:a+k],eris_OVVV,optimize=True)
del eris_OVVV
a += k
else :
eris_OVVV = radc_ao2mo.unpack_eri_1(eris.OVVV, nvir_b)
temp_singles += 0.5*lib.einsum('abc,icab->i',temp_1, eris_OVVV, optimize=True)
temp_singles -= 0.5*lib.einsum('abc,ibac->i',temp_1, eris_OVVV, optimize=True)
temp_doubles += lib.einsum('i,icab->bca',r_b,eris_OVVV,optimize=True)
temp_doubles -= lib.einsum('i,ibac->bca',r_b,eris_OVVV,optimize=True)
del eris_OVVV
s[s_b:f_b] += temp_singles
s[s_bbb:f_bbb] += 0.5*lib.einsum('bca,pbc->ap',temp_doubles,
t2_1_b_t,optimize=True).reshape(-1)
del temp_singles
del temp_doubles
r_bbb_u = np.zeros((nvir_b,nocc_b,nocc_b))
r_bbb_u[:,ij_ind_b[0],ij_ind_b[1]]= r_bbb.copy()
r_bbb_u[:,ij_ind_b[1],ij_ind_b[0]]= -r_bbb.copy()
r_bab = r_bab.reshape(nvir_b,nocc_b,nocc_a)
temp_1 = lib.einsum('jlab,ajk->blk',t2_1_b,r_bab,optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blk,lbik->i',temp_1,eris_OVoo,optimize=True)
temp = lib.einsum('jlab,ajk->blk',t2_1_b,r_bbb_u,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blk,lbik->i',temp,eris_OVOO,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blk,iblk->i',temp,eris_OVOO,optimize=True)
temp_1 = -lib.einsum('klab,akj->blj',t2_1_b,r_bab,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blj,lbij->i',temp_1,eris_OVoo,optimize=True)
temp = -lib.einsum('klab,akj->blj',t2_1_b,r_bbb_u,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blj,lbij->i',temp,eris_OVOO,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blj,iblj->i',temp,eris_OVOO,optimize=True)
temp_2 = lib.einsum('i,lbik->kbl',r_a,eris_OVoo)
temp = lib.einsum('kbl,jlab->ajk',temp_2,t2_1_b,optimize=True)
s[s_bab:f_bab] += temp.reshape(-1)
temp_1 = lib.einsum('i,lbik->kbl',r_b, eris_OVOO)
temp_1 -= lib.einsum('i,iblk->kbl',r_b, eris_OVOO)
temp = lib.einsum('kbl,jlab->ajk',temp_1,t2_1_b,optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1] ].reshape(-1)
temp_1 = lib.einsum('i,lbij->jbl',r_b, eris_OVOO)
temp_1 -= lib.einsum('i,iblj->jbl',r_b, eris_OVOO)
temp = lib.einsum('jbl,klab->ajk',temp_1,t2_1_b,optimize=True)
s[s_bbb:f_bbb] -= temp[:,ij_ind_b[0],ij_ind_b[1] ].reshape(-1)
del t2_1_b
temp_1 = lib.einsum('kjcb,ajk->abc',t2_1_ab,r_bab, optimize=True)
temp_2 = np.zeros((nvir_a, nvir_b, nvir_b))
if isinstance(eris.ovVV, type(None)):
chnk_size = uadc_ao2mo.calculate_chunk_size(adc)
a = 0
for p in range(0,nocc_a,chnk_size):
eris_ovVV = dfadc.get_ovvv_spin_df(
adc, eris.Lov, eris.LVV, p, chnk_size).reshape(-1,nvir_a,nvir_b,nvir_b)
k = eris_ovVV.shape[0]
s[s_a:f_a][a:a+k] += lib.einsum('abc,icab->i',temp_1, eris_ovVV, optimize=True)
temp_2 += lib.einsum('i,icab->cba',r_a[a:a+k],eris_ovVV,optimize=True)
del eris_ovVV
a += k
else :
eris_ovVV = radc_ao2mo.unpack_eri_1(eris.ovVV, nvir_b)
s[s_a:f_a] += lib.einsum('abc,icab->i',temp_1, eris_ovVV, optimize=True)
temp_2 += lib.einsum('i,icab->cba',r_a,eris_ovVV,optimize=True)
del eris_ovVV
s[s_bab:f_bab] += lib.einsum('cba,kjcb->ajk',temp_2, t2_1_ab, optimize=True).reshape(-1)
del temp_1
del temp_2
temp_1 = lib.einsum('jkbc,ajk->abc',t2_1_ab,r_aba, optimize=True)
temp_2 = np.zeros((nvir_a, nvir_b, nvir_a))
if isinstance(eris.OVvv, type(None)):
chnk_size = uadc_ao2mo.calculate_chunk_size(adc)
a = 0
for p in range(0,nocc_b,chnk_size):
eris_OVvv = dfadc.get_ovvv_spin_df(
adc, eris.LOV, eris.Lvv, p, chnk_size).reshape(-1,nvir_b,nvir_a,nvir_a)
k = eris_OVvv.shape[0]
s[s_b:f_b][a:a+k] += lib.einsum('abc,icab->i',temp_1, eris_OVvv, optimize=True)
temp_2 += lib.einsum('i,icab->bca',r_b[a:a+k],eris_OVvv,optimize=True)
del eris_OVvv
a += k
else :
eris_OVvv = radc_ao2mo.unpack_eri_1(eris.OVvv, nvir_a)
s[s_b:f_b] += lib.einsum('abc,icab->i',temp_1, eris_OVvv, optimize=True)
temp_2 += lib.einsum('i,icab->bca',r_b,eris_OVvv,optimize=True)
del eris_OVvv
s[s_aba:f_aba] += lib.einsum('bca,jkbc->ajk',temp_2, t2_1_ab, optimize=True).reshape(-1)
del temp_1
del temp_2
temp = lib.einsum('ljba,ajk->blk',t2_1_ab,r_bab,optimize=True)
temp_1 = lib.einsum('jlab,ajk->blk',t2_1_ab,r_aaa_u,optimize=True)
temp_2 = lib.einsum('jlba,akj->blk',t2_1_ab,r_bab, optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blk,lbik->i',temp,eris_ovoo,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blk,iblk->i',temp,eris_ovoo,optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blk,lbik->i',temp_1,eris_OVoo,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blk,iblk->i',temp_2,eris_ovOO,optimize=True)
temp = lib.einsum('jlab,ajk->blk',t2_1_ab,r_aba,optimize=True)
temp_1 = lib.einsum('ljba,ajk->blk',t2_1_ab,r_bbb_u,optimize=True)
temp_2 = lib.einsum('ljab,akj->blk',t2_1_ab,r_aba,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blk,lbik->i',temp,eris_OVOO,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blk,iblk->i',temp,eris_OVOO,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blk,lbik->i',temp_1,eris_ovOO,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blk,iblk->i',temp_2,eris_OVoo,optimize=True)
temp = -lib.einsum('lkba,akj->blj',t2_1_ab,r_bab,optimize=True)
temp_1 = -lib.einsum('klab,akj->blj',t2_1_ab,r_aaa_u,optimize=True)
temp_2 = -lib.einsum('klba,ajk->blj',t2_1_ab,r_bab,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blj,lbij->i',temp,eris_ovoo,optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blj,iblj->i',temp,eris_ovoo,optimize=True)
s[s_a:f_a] -= 0.5*lib.einsum('blj,lbij->i',temp_1,eris_OVoo,optimize=True)
s[s_a:f_a] += 0.5*lib.einsum('blj,iblj->i',temp_2,eris_ovOO,optimize=True)
temp = -lib.einsum('klab,akj->blj',t2_1_ab,r_aba,optimize=True)
temp_1 = -lib.einsum('lkba,akj->blj',t2_1_ab,r_bbb_u,optimize=True)
temp_2 = -lib.einsum('lkab,ajk->blj',t2_1_ab,r_aba,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blj,lbij->i',temp,eris_OVOO,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blj,iblj->i',temp,eris_OVOO,optimize=True)
s[s_b:f_b] -= 0.5*lib.einsum('blj,lbij->i',temp_1,eris_ovOO,optimize=True)
s[s_b:f_b] += 0.5*lib.einsum('blj,iblj->i',temp_2,eris_OVoo,optimize=True)
temp_2 = lib.einsum('i,lbik->kbl',r_a, eris_OVoo)
temp = lib.einsum('kbl,jlab->ajk',temp_2,t2_1_ab,optimize=True)
s[s_aaa:f_aaa] += temp[:,ij_ind_a[0],ij_ind_a[1] ].reshape(-1)
temp_1 = lib.einsum('i,lbik->kbl',r_a,eris_ovoo)
temp_1 -= lib.einsum('i,iblk->kbl',r_a,eris_ovoo)
temp = lib.einsum('kbl,ljba->ajk',temp_1,t2_1_ab,optimize=True)
s[s_bab:f_bab] += temp.reshape(-1)
temp_2 = lib.einsum('i,lbik->kbl',r_b, eris_ovOO)
temp = lib.einsum('kbl,ljba->ajk',temp_2,t2_1_ab,optimize=True)
s[s_bbb:f_bbb] += temp[:,ij_ind_b[0],ij_ind_b[1] ].reshape(-1)
temp_1 = lib.einsum('i,lbik->kbl',r_b,eris_OVOO)
temp_1 -= lib.einsum('i,iblk->kbl',r_b,eris_OVOO)
temp = lib.einsum('kbl,jlab->ajk',temp_1,t2_1_ab,optimize=True)
s[s_aba:f_aba] += temp.reshape(-1)
temp_2 = lib.einsum('i,lbij->jbl',r_a, eris_OVoo)
temp = lib.einsum('jbl,klab->ajk',temp_2,t2_1_ab,optimize=True)
s[s_aaa:f_aaa] -= temp[:,ij_ind_a[0],ij_ind_a[1] ].reshape(-1)
temp = -lib.einsum('i,iblj->jbl',r_a,eris_ovOO,optimize=True)
temp_1 = -lib.einsum('jbl,klba->ajk',temp,t2_1_ab,optimize=True)
s[s_bab:f_bab] -= temp_1.reshape(-1)
temp_2 = lib.einsum('i,lbij->jbl',r_b, eris_ovOO)
temp = lib.einsum('jbl,lkba->ajk',temp_2,t2_1_ab,optimize=True)
s[s_bbb:f_bbb] -= temp[:,ij_ind_b[0],ij_ind_b[1] ].reshape(-1)
temp = -lib.einsum('i,iblj->jbl',r_b,eris_OVoo,optimize=True)
temp_1 = -lib.einsum('jbl,lkab->ajk',temp,t2_1_ab,optimize=True)
s[s_aba:f_aba] -= temp_1.reshape(-1)
del t2_1_ab
cput0 = log.timer_debug1("completed sigma vector calculation", *cput0)
s *= -1.0
return s
return sigma_
[docs]
def get_trans_moments(adc):
cput0 = (logger.process_clock(), logger.perf_counter())
log = logger.Logger(adc.stdout, adc.verbose)
nmo_a = adc.nmo_a
nmo_b = adc.nmo_b
T_a = []
T_b = []
for orb in range(nmo_a):
T_aa = get_trans_moments_orbital(adc,orb, spin="alpha")
T_a.append(T_aa)
for orb in range(nmo_b):
T_bb = get_trans_moments_orbital(adc,orb, spin="beta")
T_b.append(T_bb)
cput0 = log.timer_debug1("completed spec vector calc in ADC(3) calculation", *cput0)
return (T_a, T_b)
[docs]
def get_trans_moments_orbital(adc, orb, spin="alpha"):
if adc.method not in ("adc(2)", "adc(2)-x", "adc(3)"):
raise NotImplementedError(adc.method)
method = adc.method
if (adc.approx_trans_moments is False or adc.method == "adc(3)"):
t1_2_a, t1_2_b = adc.t1[0]
t2_1_a = adc.t2[0][0][:]
t2_1_ab = adc.t2[0][1][:]
t2_1_b = adc.t2[0][2][:]
nocc_a = adc.nocc_a
nocc_b = adc.nocc_b
nvir_a = adc.nvir_a
nvir_b = adc.nvir_b
ij_ind_a = np.tril_indices(nocc_a, k=-1)
ij_ind_b = np.tril_indices(nocc_b, k=-1)
n_singles_a = nocc_a
n_singles_b = nocc_b
n_doubles_aaa = nocc_a* (nocc_a - 1) * nvir_a // 2
n_doubles_bab = nvir_b * nocc_a* nocc_b
n_doubles_aba = nvir_a * nocc_b* nocc_a
n_doubles_bbb = nocc_b* (nocc_b - 1) * nvir_b // 2
dim = n_singles_a + n_singles_b + n_doubles_aaa + n_doubles_bab + n_doubles_aba + n_doubles_bbb
idn_occ_a = np.identity(nocc_a)
idn_occ_b = np.identity(nocc_b)
s_a = 0
f_a = n_singles_a
s_b = f_a
f_b = s_b + n_singles_b
s_aaa = f_b
f_aaa = s_aaa + n_doubles_aaa
s_bab = f_aaa
f_bab = s_bab + n_doubles_bab
s_aba = f_bab
f_aba = s_aba + n_doubles_aba
s_bbb = f_aba
f_bbb = s_bbb + n_doubles_bbb
T = np.zeros((dim))
######## spin = alpha ############################################
if spin == "alpha":
pass # placeholder to mute flake8 warning
######## ADC(2) 1h part ############################################
t2_1_a = adc.t2[0][0][:]
t2_1_ab = adc.t2[0][1][:]
if orb < nocc_a:
T[s_a:f_a] = idn_occ_a[orb, :]
T[s_a:f_a] += 0.25*lib.einsum('kdc,ikdc->i',t2_1_a[:,orb,:,:], t2_1_a, optimize=True)
T[s_a:f_a] -= 0.25*lib.einsum('kdc,ikdc->i',t2_1_ab[orb,:,:,:], t2_1_ab, optimize=True)
T[s_a:f_a] -= 0.25*lib.einsum('kcd,ikcd->i',t2_1_ab[orb,:,:,:], t2_1_ab, optimize=True)
else:
if (adc.approx_trans_moments is False or adc.method == "adc(3)"):
T[s_a:f_a] += t1_2_a[:,(orb-nocc_a)]
######## ADC(2) 2h-1p part ############################################
t2_1_t = t2_1_a[ij_ind_a[0],ij_ind_a[1],:,:]
t2_1_t_a = t2_1_t.transpose(2,1,0)
t2_1_t_ab = t2_1_ab.transpose(2,3,1,0)
T[s_aaa:f_aaa] = t2_1_t_a[(orb-nocc_a),:,:].reshape(-1)
T[s_bab:f_bab] = t2_1_t_ab[(orb-nocc_a),:,:,:].reshape(-1)
######## ADC(3) 2h-1p part ############################################
if (adc.method == "adc(2)-x" and adc.approx_trans_moments is False) or (adc.method == "adc(3)"):
t2_2_a = adc.t2[1][0][:]
t2_2_ab = adc.t2[1][1][:]
if orb >= nocc_a:
t2_2_t = t2_2_a[ij_ind_a[0],ij_ind_a[1],:,:]
t2_2_t_a = t2_2_t.transpose(2,1,0)
t2_2_t_ab = t2_2_ab.transpose(2,3,1,0)
T[s_aaa:f_aaa] += t2_2_t_a[(orb-nocc_a),:,:].reshape(-1)
T[s_bab:f_bab] += t2_2_t_ab[(orb-nocc_a),:,:,:].reshape(-1)
######## ADC(3) 1h part ############################################
if (method == 'adc(3)'):
if (adc.approx_trans_moments is False):
t1_3_a, t1_3_b = adc.t1[1]
if orb < nocc_a:
t2_1_a_tmp = np.ascontiguousarray(t2_1_a[:,orb,:,:])
t2_1_ab_tmp = np.ascontiguousarray(t2_1_ab[orb,:,:,:])
T[s_a:f_a] += 0.25*lib.einsum('kdc,ikdc->i',t2_1_a_tmp, t2_2_a, optimize=True)
T[s_a:f_a] -= 0.25*lib.einsum('kdc,ikdc->i',t2_1_ab_tmp, t2_2_ab, optimize=True)
T[s_a:f_a] -= 0.25*lib.einsum('kcd,ikcd->i',t2_1_ab_tmp, t2_2_ab, optimize=True)
del t2_1_a_tmp, t2_1_ab_tmp
t2_2_a_tmp = np.ascontiguousarray(t2_2_a[:,orb,:,:])
t2_2_ab_tmp = np.ascontiguousarray(t2_2_ab[orb,:,:,:])
T[s_a:f_a] += 0.25*lib.einsum('ikdc,kdc->i',t2_1_a, t2_2_a_tmp,optimize=True)
T[s_a:f_a] -= 0.25*lib.einsum('ikcd,kcd->i',t2_1_ab, t2_2_ab_tmp,optimize=True)
T[s_a:f_a] -= 0.25*lib.einsum('ikdc,kdc->i',t2_1_ab, t2_2_ab_tmp,optimize=True)
del t2_2_a_tmp, t2_2_ab_tmp
else:
t2_1_a_tmp = np.ascontiguousarray(t2_1_a[:,:,(orb-nocc_a),:])
t2_1_ab_tmp = np.ascontiguousarray(t2_1_ab[:,:,(orb-nocc_a),:])
T[s_a:f_a] += 0.5*lib.einsum('ikc,kc->i',t2_1_a_tmp, t1_2_a,optimize=True)
T[s_a:f_a] += 0.5*lib.einsum('ikc,kc->i',t2_1_ab_tmp, t1_2_b,optimize=True)
if (adc.approx_trans_moments is False):
T[s_a:f_a] += t1_3_a[:,(orb-nocc_a)]
del t2_1_a_tmp, t2_1_ab_tmp
del t2_2_a
del t2_2_ab
del t2_1_a
del t2_1_ab
######## spin = beta ############################################
else:
pass # placeholder
######## ADC(2) 1h part ############################################
t2_1_b = adc.t2[0][2][:]
t2_1_ab = adc.t2[0][1][:]
if orb < nocc_b:
t2_1_b_tmp = np.ascontiguousarray(t2_1_b[:,orb,:,:])
t2_1_ab_tmp = np.ascontiguousarray(t2_1_ab[:,orb,:,:])
T[s_b:f_b] = idn_occ_b[orb, :]
T[s_b:f_b]+= 0.25*lib.einsum('kdc,ikdc->i',t2_1_b_tmp, t2_1_b, optimize=True)
T[s_b:f_b]-= 0.25*lib.einsum('kdc,kidc->i',t2_1_ab_tmp, t2_1_ab, optimize=True)
T[s_b:f_b]-= 0.25*lib.einsum('kcd,kicd->i',t2_1_ab_tmp, t2_1_ab, optimize=True)
del t2_1_b_tmp, t2_1_ab_tmp
else:
if (adc.approx_trans_moments is False or adc.method == "adc(3)"):
T[s_b:f_b] += t1_2_b[:,(orb-nocc_b)]
######## ADC(2) 2h-1p part ############################################
t2_1_t = t2_1_b[ij_ind_b[0],ij_ind_b[1],:,:]
t2_1_t_b = t2_1_t.transpose(2,1,0)
t2_1_t_ab = t2_1_ab.transpose(2,3,0,1)
T[s_bbb:f_bbb] = t2_1_t_b[(orb-nocc_b),:,:].reshape(-1)
T[s_aba:f_aba] = t2_1_t_ab[:,(orb-nocc_b),:,:].reshape(-1)
######## ADC(3) 2h-1p part ############################################
if (adc.method == "adc(2)-x" and adc.approx_trans_moments is False) or (adc.method == "adc(3)"):
t2_2_a = adc.t2[1][0][:]
t2_2_ab = adc.t2[1][1][:]
t2_2_b = adc.t2[1][2][:]
if orb >= nocc_b:
t2_2_t = t2_2_b[ij_ind_b[0],ij_ind_b[1],:,:]
t2_2_t_b = t2_2_t.transpose(2,1,0)
t2_2_t_ab = t2_2_ab.transpose(2,3,0,1)
T[s_bbb:f_bbb] += t2_2_t_b[(orb-nocc_b),:,:].reshape(-1)
T[s_aba:f_aba] += t2_2_t_ab[:,(orb-nocc_b),:,:].reshape(-1)
######## ADC(3) 1h part ############################################
if (method=='adc(3)'):
if (adc.approx_trans_moments is False):
t1_3_a, t1_3_b = adc.t1[1]
if orb < nocc_b:
t2_1_b_tmp = np.ascontiguousarray(t2_1_b[:,orb,:,:])
t2_1_ab_tmp = np.ascontiguousarray(t2_1_ab[:,orb,:,:])
T[s_b:f_b] += 0.25*lib.einsum('kdc,ikdc->i',t2_1_b_tmp, t2_2_b, optimize=True)
T[s_b:f_b] -= 0.25*lib.einsum('kdc,kidc->i',t2_1_ab_tmp, t2_2_ab, optimize=True)
T[s_b:f_b] -= 0.25*lib.einsum('kcd,kicd->i',t2_1_ab_tmp, t2_2_ab, optimize=True)
del t2_1_b_tmp, t2_1_ab_tmp
t2_2_b_tmp = np.ascontiguousarray(t2_2_b[:,orb,:,:])
t2_2_ab_tmp = np.ascontiguousarray(t2_2_ab[:,orb,:,:])
T[s_b:f_b] += 0.25*lib.einsum('ikdc,kdc->i',t2_1_b, t2_2_b_tmp ,optimize=True)
T[s_b:f_b] -= 0.25*lib.einsum('kicd,kcd->i',t2_1_ab, t2_2_ab_tmp,optimize=True)
T[s_b:f_b] -= 0.25*lib.einsum('kidc,kdc->i',t2_1_ab, t2_2_ab_tmp,optimize=True)
del t2_2_b_tmp, t2_2_ab_tmp
else:
t2_1_b_tmp = np.ascontiguousarray(t2_1_b[:,:,(orb-nocc_b),:])
t2_1_ab_tmp = np.ascontiguousarray(t2_1_ab[:,:,:,(orb-nocc_b)])
T[s_b:f_b] += 0.5*lib.einsum('ikc,kc->i',t2_1_b_tmp, t1_2_b,optimize=True)
T[s_b:f_b] += 0.5*lib.einsum('kic,kc->i',t2_1_ab_tmp, t1_2_a,optimize=True)
if (adc.approx_trans_moments is False):
T[s_b:f_b] += t1_3_b[:,(orb-nocc_b)]
del t2_1_b_tmp, t2_1_ab_tmp
del t2_2_b
del t2_2_ab
del t2_1_b
del t2_1_ab
return T
[docs]
def analyze_eigenvector(adc):
nocc_a = adc.nocc_a
nocc_b = adc.nocc_b
nvir_a = adc.nvir_a
nvir_b = adc.nvir_b
evec_print_tol = adc.evec_print_tol
logger.info(adc, "Number of alpha occupied orbitals = %d", nocc_a)
logger.info(adc, "Number of beta occupied orbitals = %d", nocc_b)
logger.info(adc, "Number of alpha virtual orbitals = %d", nvir_a)
logger.info(adc, "Number of beta virtual orbitals = %d", nvir_b)
logger.info(adc, "Print eigenvector elements > %f\n", evec_print_tol)
ij_a = np.tril_indices(nocc_a, k=-1)
ij_b = np.tril_indices(nocc_b, k=-1)
n_singles_a = nocc_a
n_singles_b = nocc_b
n_doubles_aaa = nocc_a* (nocc_a - 1) * nvir_a // 2
n_doubles_bab = nvir_b * nocc_a* nocc_b
n_doubles_aba = nvir_a * nocc_b* nocc_a
n_doubles_bbb = nocc_b* (nocc_b - 1) * nvir_b // 2
s_a = 0
f_a = n_singles_a
s_b = f_a
f_b = s_b + n_singles_b
s_aaa = f_b
f_aaa = s_aaa + n_doubles_aaa
s_bab = f_aaa
f_bab = s_bab + n_doubles_bab
s_aba = f_bab
f_aba = s_aba + n_doubles_aba
s_bbb = f_aba
f_bbb = s_bbb + n_doubles_bbb
U = adc.U
for I in range(U.shape[1]):
U1 = U[:f_b,I]
U2 = U[f_b:,I]
U1dotU1 = np.dot(U1, U1)
U2dotU2 = np.dot(U2, U2)
temp_aaa = np.zeros((nvir_a, nocc_a, nocc_a))
temp_aaa[:,ij_a[0],ij_a[1]] = U[s_aaa:f_aaa,I].reshape(nvir_a,-1).copy()
temp_aaa[:,ij_a[1],ij_a[0]] = -U[s_aaa:f_aaa,I].reshape(nvir_a,-1).copy()
U_aaa = temp_aaa.reshape(-1).copy()
temp_bbb = np.zeros((nvir_b, nocc_b, nocc_b))
temp_bbb[:,ij_b[0],ij_b[1]] = U[s_bbb:f_bbb,I].reshape(nvir_b,-1).copy()
temp_bbb[:,ij_b[1],ij_b[0]] = -U[s_bbb:f_bbb,I].reshape(nvir_b,-1).copy()
U_bbb = temp_bbb.reshape(-1).copy()
U_sq = U[:,I].copy()**2
ind_idx = np.argsort(-U_sq)
U_sq = U_sq[ind_idx]
U_sorted = U[ind_idx,I].copy()
U_sq_aaa = U_aaa.copy()**2
U_sq_bbb = U_bbb.copy()**2
ind_idx_aaa = np.argsort(-U_sq_aaa)
ind_idx_bbb = np.argsort(-U_sq_bbb)
U_sq_aaa = U_sq_aaa[ind_idx_aaa]
U_sq_bbb = U_sq_bbb[ind_idx_bbb]
U_sorted_aaa = U_aaa[ind_idx_aaa].copy()
U_sorted_bbb = U_bbb[ind_idx_bbb].copy()
U_sorted = U_sorted[U_sq > evec_print_tol**2]
ind_idx = ind_idx[U_sq > evec_print_tol**2]
U_sorted_aaa = U_sorted_aaa[U_sq_aaa > evec_print_tol**2]
U_sorted_bbb = U_sorted_bbb[U_sq_bbb > evec_print_tol**2]
ind_idx_aaa = ind_idx_aaa[U_sq_aaa > evec_print_tol**2]
ind_idx_bbb = ind_idx_bbb[U_sq_bbb > evec_print_tol**2]
singles_a_idx = []
singles_b_idx = []
doubles_aaa_idx = []
doubles_bab_idx = []
doubles_aba_idx = []
doubles_bbb_idx = []
singles_a_val = []
singles_b_val = []
doubles_bab_val = []
doubles_aba_val = []
iter_idx = 0
for orb_idx in ind_idx:
if orb_idx in range(s_a,f_a):
i_idx = orb_idx + 1
singles_a_idx.append(i_idx)
singles_a_val.append(U_sorted[iter_idx])
if orb_idx in range(s_b,f_b):
i_idx = orb_idx - s_b + 1
singles_b_idx.append(i_idx)
singles_b_val.append(U_sorted[iter_idx])
if orb_idx in range(s_bab,f_bab):
aij_idx = orb_idx - s_bab
ij_rem = aij_idx % (nocc_a*nocc_b)
a_idx = aij_idx//(nocc_a*nocc_b)
i_idx = ij_rem//nocc_a
j_idx = ij_rem % nocc_a
doubles_bab_idx.append((a_idx + 1 + nocc_b, i_idx + 1, j_idx + 1))
doubles_bab_val.append(U_sorted[iter_idx])
if orb_idx in range(s_aba,f_aba):
aij_idx = orb_idx - s_aba
ij_rem = aij_idx % (nocc_b*nocc_a)
a_idx = aij_idx//(nocc_b*nocc_a)
i_idx = ij_rem//nocc_b
j_idx = ij_rem % nocc_b
doubles_aba_idx.append((a_idx + 1 + nocc_a, i_idx + 1, j_idx + 1))
doubles_aba_val.append(U_sorted[iter_idx])
iter_idx += 1
for orb_aaa in ind_idx_aaa:
ij_rem = orb_aaa % (nocc_a*nocc_a)
a_idx = orb_aaa//(nocc_a*nocc_a)
i_idx = ij_rem//nocc_a
j_idx = ij_rem % nocc_a
doubles_aaa_idx.append((a_idx + 1 + nocc_a, i_idx + 1, j_idx + 1))
for orb_bbb in ind_idx_bbb:
ij_rem = orb_bbb % (nocc_b*nocc_b)
a_idx = orb_bbb//(nocc_b*nocc_b)
i_idx = ij_rem//nocc_b
j_idx = ij_rem % nocc_b
doubles_bbb_idx.append((a_idx + 1 + nocc_b, i_idx + 1, j_idx + 1))
doubles_aaa_val = list(U_sorted_aaa)
doubles_bbb_val = list(U_sorted_bbb)
logger.info(adc,'%s | root %d | norm(1h) = %6.4f | norm(2h1p) = %6.4f ',
adc.method ,I, U1dotU1, U2dotU2)
if singles_a_val:
logger.info(adc, "\n1h(alpha) block: ")
logger.info(adc, " i U(i)")
logger.info(adc, "------------------")
for idx, print_singles in enumerate(singles_a_idx):
logger.info(adc, ' %4d %7.4f', print_singles, singles_a_val[idx])
if singles_b_val:
logger.info(adc, "\n1h(beta) block: ")
logger.info(adc, " i U(i)")
logger.info(adc, "------------------")
for idx, print_singles in enumerate(singles_b_idx):
logger.info(adc, ' %4d %7.4f', print_singles, singles_b_val[idx])
if doubles_aaa_val:
logger.info(adc, "\n2h1p(alpha|alpha|alpha) block: ")
logger.info(adc, " i j a U(i,j,a)")
logger.info(adc, "-------------------------------")
for idx, print_doubles in enumerate(doubles_aaa_idx):
logger.info(adc, ' %4d %4d %4d %7.4f',
print_doubles[1], print_doubles[2], print_doubles[0], doubles_aaa_val[idx])
if doubles_bab_val:
logger.info(adc, "\n2h1p(beta|alpha|beta) block: ")
logger.info(adc, " i j a U(i,j,a)")
logger.info(adc, "-------------------------------")
for idx, print_doubles in enumerate(doubles_bab_idx):
logger.info(adc, ' %4d %4d %4d %7.4f', print_doubles[1],
print_doubles[2], print_doubles[0], doubles_bab_val[idx])
if doubles_aba_val:
logger.info(adc, "\n2h1p(alpha|beta|alpha) block: ")
logger.info(adc, " i j a U(i,j,a)")
logger.info(adc, "-------------------------------")
for idx, print_doubles in enumerate(doubles_aba_idx):
logger.info(adc, ' %4d %4d %4d %7.4f',
print_doubles[1], print_doubles[2], print_doubles[0], doubles_aba_val[idx])
if doubles_bbb_val:
logger.info(adc, "\n2h1p(beta|beta|beta) block: ")
logger.info(adc, " i j a U(i,j,a)")
logger.info(adc, "-------------------------------")
for idx, print_doubles in enumerate(doubles_bbb_idx):
logger.info(adc, ' %4d %4d %4d %7.4f',
print_doubles[1], print_doubles[2], print_doubles[0], doubles_bbb_val[idx])
logger.info(adc, "\n*************************************************************\n")
[docs]
def analyze_spec_factor(adc):
X_a = adc.X[0]
X_b = adc.X[1]
logger.info(adc, "Print spectroscopic factors > %E\n", adc.spec_factor_print_tol)
X_tot = (X_a, X_b)
for iter_idx, X in enumerate(X_tot):
if iter_idx == 0:
spin = "alpha"
else:
spin = "beta"
X_2 = (X.copy()**2)
thresh = adc.spec_factor_print_tol
for i in range(X_2.shape[1]):
sort = np.argsort(-X_2[:,i])
X_2_row = X_2[:,i]
X_2_row = X_2_row[sort]
if not adc.mol.symmetry:
sym = np.repeat(['A'], X_2_row.shape[0])
else:
if spin == "alpha":
sym = [symm.irrep_id2name(adc.mol.groupname, x)
for x in adc._scf.mo_coeff[0].orbsym]
sym = np.array(sym)
else:
sym = [symm.irrep_id2name(adc.mol.groupname, x)
for x in adc._scf.mo_coeff[1].orbsym]
sym = np.array(sym)
sym = sym[sort]
spec_Contribution = X_2_row[X_2_row > thresh]
index_mo = sort[X_2_row > thresh]+1
if np.sum(spec_Contribution) == 0.0:
continue
logger.info(adc, '%s | root %d %s\n', adc.method, i, spin)
logger.info(adc, " HF MO Spec. Contribution Orbital symmetry")
logger.info(adc, "-----------------------------------------------------------")
for c in range(index_mo.shape[0]):
logger.info(adc, ' %3.d %10.8f %s',
index_mo[c], spec_Contribution[c], sym[c])
logger.info(adc, '\nPartial spec. factor sum = %10.8f', np.sum(spec_Contribution))
logger.info(adc, "\n*************************************************************\n")
[docs]
def get_properties(adc, nroots=1):
#Transition moments
T = adc.get_trans_moments()
T_a = T[0]
T_b = T[1]
T_a = np.array(T_a)
T_b = np.array(T_b)
U = adc.U
#Spectroscopic amplitudes
X_a = np.dot(T_a, U).reshape(-1,nroots)
X_b = np.dot(T_b, U).reshape(-1,nroots)
X = (X_a,X_b)
#Spectroscopic factors
P = lib.einsum("pi,pi->i", X_a, X_a)
P += lib.einsum("pi,pi->i", X_b, X_b)
return P, X
[docs]
def analyze(myadc):
header = ("\n*************************************************************"
"\n Eigenvector analysis summary"
"\n*************************************************************")
logger.info(myadc, header)
myadc.analyze_eigenvector()
if myadc.compute_properties:
header = ("\n*************************************************************"
"\n Spectroscopic factors analysis summary"
"\n*************************************************************")
logger.info(myadc, header)
myadc.analyze_spec_factor()
[docs]
def compute_dyson_mo(myadc):
X_a = myadc.X[0]
X_b = myadc.X[1]
if X_a is None:
nroots = myadc.U.shape[1]
P,X_a,X_b = myadc.get_properties(nroots)
nroots = X_a.shape[1]
dyson_mo_a = np.dot(myadc.mo_coeff[0],X_a)
dyson_mo_b = np.dot(myadc.mo_coeff[1],X_b)
dyson_mo = (dyson_mo_a,dyson_mo_b)
return dyson_mo
[docs]
def make_rdm1(adc):
cput0 = (logger.process_clock(), logger.perf_counter())
log = logger.Logger(adc.stdout, adc.verbose)
U = adc.U
list_rdm1_a = []
list_rdm1_b = []
for i in range(U.shape[1]):
rdm1_a, rdm1_b = make_rdm1_eigenvectors(adc, U[:,i], U[:,i])
list_rdm1_a.append(rdm1_a)
list_rdm1_b.append(rdm1_b)
cput0 = log.timer_debug1("completed OPDM calculation", *cput0)
return (list_rdm1_a, list_rdm1_b)
[docs]
def make_rdm1_eigenvectors(adc, L, R):
L = np.array(L).ravel()
R = np.array(R).ravel()
t2_1_a = adc.t2[0][0][:]
t2_1_ab = adc.t2[0][1][:]
t2_1_b = adc.t2[0][2][:]
t1_2_a = adc.t1[0][0][:]
t1_2_b = adc.t1[0][1][:]
nocc_a = adc.nocc_a
nocc_b = adc.nocc_b
nvir_a = adc.nvir_a
nvir_b = adc.nvir_b
nmo_a = nocc_a + nvir_a
nmo_b = nocc_b + nvir_b
ij_ind_a = np.tril_indices(nocc_a, k=-1)
ij_ind_b = np.tril_indices(nocc_b, k=-1)
n_singles_a = nocc_a
n_singles_b = nocc_b
n_doubles_aaa = nocc_a * (nocc_a - 1) * nvir_a // 2
n_doubles_bab = nvir_b * nocc_a * nocc_b
n_doubles_aba = nvir_a * nocc_b * nocc_a
n_doubles_bbb = nocc_b * (nocc_b - 1) * nvir_b // 2
s_a = 0
f_a = n_singles_a
s_b = f_a
f_b = s_b + n_singles_b
s_aaa = f_b
f_aaa = s_aaa + n_doubles_aaa
s_bab = f_aaa
f_bab = s_bab + n_doubles_bab
s_aba = f_bab
f_aba = s_aba + n_doubles_aba
s_bbb = f_aba
f_bbb = s_bbb + n_doubles_bbb
rdm1_a = np.zeros((nmo_a,nmo_a))
rdm1_b = np.zeros((nmo_b,nmo_b))
kd_oc_a = np.identity(nocc_a)
kd_oc_b = np.identity(nocc_b)
L_a = L[s_a:f_a]
L_b = L[s_b:f_b]
L_aaa = L[s_aaa:f_aaa]
L_bab = L[s_bab:f_bab]
L_aba = L[s_aba:f_aba]
L_bbb = L[s_bbb:f_bbb]
R_a = R[s_a:f_a]
R_b = R[s_b:f_b]
R_aaa = R[s_aaa:f_aaa]
R_bab = R[s_bab:f_bab]
R_aba = R[s_aba:f_aba]
R_bbb = R[s_bbb:f_bbb]
L_aaa = L_aaa.reshape(nvir_a,-1)
L_bbb = L_bbb.reshape(nvir_b,-1)
L_aaa_u = None
L_aaa_u = np.zeros((nvir_a,nocc_a,nocc_a))
L_aaa_u[:,ij_ind_a[0],ij_ind_a[1]]= L_aaa.copy()
L_aaa_u[:,ij_ind_a[1],ij_ind_a[0]]= -L_aaa.copy()
L_bbb_u = None
L_bbb_u = np.zeros((nvir_b,nocc_b,nocc_b))
L_bbb_u[:,ij_ind_b[0],ij_ind_b[1]]= L_bbb.copy()
L_bbb_u[:,ij_ind_b[1],ij_ind_b[0]]= -L_bbb.copy()
L_aba = L_aba.reshape(nvir_a,nocc_a,nocc_b)
L_bab = L_bab.reshape(nvir_b,nocc_b,nocc_a)
R_aaa = R_aaa.reshape(nvir_a,-1)
R_bbb = R_bbb.reshape(nvir_b,-1)
R_aaa_u = None
R_aaa_u = np.zeros((nvir_a,nocc_a,nocc_a))
R_aaa_u[:,ij_ind_a[0],ij_ind_a[1]]= R_aaa.copy()
R_aaa_u[:,ij_ind_a[1],ij_ind_a[0]]= -R_aaa.copy()
R_bbb_u = None
R_bbb_u = np.zeros((nvir_b,nocc_b,nocc_b))
R_bbb_u[:,ij_ind_b[0],ij_ind_b[1]]= R_bbb.copy()
R_bbb_u[:,ij_ind_b[1],ij_ind_b[0]]= -R_bbb.copy()
R_aba = R_aba.reshape(nvir_a,nocc_a,nocc_b)
R_bab = R_bab.reshape(nvir_b,nocc_b,nocc_a)
######### block- ij
rdm1_a[:nocc_a,:nocc_a] = np.einsum('ij,m,m->ij',kd_oc_a,L_a,R_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= np.einsum('i,j->ij',L_a,R_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += np.einsum('ij,m,m->ij',kd_oc_a,L_b,R_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] = np.einsum('ij,m,m->ij',kd_oc_b,L_b,R_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= np.einsum('i,j->ij',L_b,R_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += np.einsum('ij,m,m->ij',kd_oc_b,L_a,R_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.5*np.einsum('ij,etu,etu->ij',kd_oc_a,L_aaa_u,R_aaa_u,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += np.einsum('ij,etu,etu->ij',kd_oc_a,L_bab,R_bab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += np.einsum('ij,etu,etu->ij',kd_oc_a,L_aba,R_aba,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.5*np.einsum('ij,etu,etu->ij',kd_oc_a,L_bbb_u,R_bbb_u,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= np.einsum('eti,etj->ij',L_aaa_u,R_aaa_u,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= np.einsum('eti,etj->ij',L_bab,R_bab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= np.einsum('eit,ejt->ij',L_aba,R_aba,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.5*np.einsum('ij,etu,etu->ij',kd_oc_b,L_aaa_u,R_aaa_u,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += np.einsum('ij,etu,etu->ij',kd_oc_b,L_bab,R_bab,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += np.einsum('ij,etu,etu->ij',kd_oc_b,L_aba,R_aba,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.5*np.einsum('ij,etu,etu->ij',kd_oc_b,L_bbb_u,R_bbb_u,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= np.einsum('eti,etj->ij',L_bbb_u,R_bbb_u,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= np.einsum('eti,etj->ij',L_aba,R_aba,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= np.einsum('eit,ejt->ij',L_bab,R_bab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= 0.5* \
np.einsum('g,g,hjcd,hicd->ij', L_a,R_a,t2_1_a,t2_1_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= 0.5* \
np.einsum('g,g,hjcd,hicd->ij', L_b,R_b,t2_1_a,t2_1_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= np.einsum('g,g,jhcd,ihcd->ij', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] -= np.einsum('g,g,jhcd,ihcd->ij', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.5* \
np.einsum('g,h,jgcd,ihcd->ij', L_a,R_a,t2_1_a,t2_1_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += np.einsum('g,h,jgcd,ihcd->ij', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.25* \
np.einsum('g,j,ghcd,ihcd->ij',L_a,R_a,t2_1_a,t2_1_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.5* \
np.einsum('g,j,ghcd,ihcd->ij',L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.25* \
np.einsum('g,i,jhcd,ghcd->ij',R_a,L_a,t2_1_a,t2_1_a,optimize=True)
rdm1_a[:nocc_a,:nocc_a] += 0.5* \
np.einsum('g,i,jhcd,ghcd->ij',R_a,L_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= 0.5* \
np.einsum('g,g,hjcd,hicd->ij', L_b,R_b,t2_1_b,t2_1_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= 0.5* \
np.einsum('g,g,hjcd,hicd->ij', L_a,R_a,t2_1_b,t2_1_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= np.einsum('g,g,hjcd,hicd->ij', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,:nocc_b] -= np.einsum('g,g,hjcd,hicd->ij', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.5* \
np.einsum('g,h,jgcd,ihcd->ij', L_b,R_b,t2_1_b,t2_1_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += np.einsum('g,h,gjcd,hicd->ij', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.25* \
np.einsum('g,j,ghcd,ihcd->ij',L_b,R_b,t2_1_b,t2_1_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.5* \
np.einsum('g,j,hgcd,hicd->ij',L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.25* \
np.einsum('g,i,jhcd,ghcd->ij',R_b,L_b,t2_1_b,t2_1_b,optimize=True)
rdm1_b[:nocc_b,:nocc_b] += 0.5* \
np.einsum('g,i,hjcd,hgcd->ij',R_b,L_b,t2_1_ab,t2_1_ab,optimize=True)
########## block- ab
rdm1_a[nocc_a:,nocc_a:] = 0.5*np.einsum('atu,btu->ab', L_aaa_u,R_aaa_u,optimize=True)
rdm1_a[nocc_a:,nocc_a:] += np.einsum('atu,btu->ab', L_aba,R_aba,optimize=True)
rdm1_b[nocc_b:,nocc_b:] = 0.5*np.einsum('atu,btu->ab', L_bbb_u,R_bbb_u,optimize=True)
rdm1_b[nocc_b:,nocc_b:] += np.einsum('atu,btu->ab', L_bab,R_bab,optimize=True)
rdm1_a[nocc_a:,nocc_a:] += 0.5* \
np.einsum('g,g,hmbc,hmac->ab', L_a,R_a,t2_1_a,t2_1_a,optimize=True)
rdm1_a[nocc_a:,nocc_a:] += 0.5* \
np.einsum('g,g,hmbc,hmac->ab', L_b,R_b,t2_1_a,t2_1_a,optimize=True)
rdm1_a[nocc_a:,nocc_a:] += np.einsum('g,g,hmbc,hmac->ab', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[nocc_a:,nocc_a:] += np.einsum('g,g,hmbc,hmac->ab', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[nocc_a:,nocc_a:] -= np.einsum('g,h,hmbc,gmac->ab', L_a,R_a,t2_1_a,t2_1_a,optimize=True)
rdm1_a[nocc_a:,nocc_a:] -= np.einsum('g,h,hmbc,gmac->ab', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_a[nocc_a:,nocc_a:] -= np.einsum('g,h,mhbc,mgac->ab', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[nocc_b:,nocc_b:] += 0.5* \
np.einsum('g,g,hmbc,hmac->ab', L_b,R_b,t2_1_b,t2_1_b,optimize=True)
rdm1_b[nocc_b:,nocc_b:] += 0.5* \
np.einsum('g,g,hmbc,hmac->ab', L_a,R_a,t2_1_b,t2_1_b,optimize=True)
rdm1_b[nocc_b:,nocc_b:] += np.einsum('g,g,hmcb,hmca->ab', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[nocc_b:,nocc_b:] += np.einsum('g,g,hmcb,hmca->ab', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[nocc_b:,nocc_b:] -= np.einsum('g,h,hmbc,gmac->ab', L_b,R_b,t2_1_b,t2_1_b,optimize=True)
rdm1_b[nocc_b:,nocc_b:] -= np.einsum('g,h,mhcb,mgca->ab', L_b,R_b,t2_1_ab,t2_1_ab,optimize=True)
rdm1_b[nocc_b:,nocc_b:] -= np.einsum('g,h,hmcb,gmca->ab', L_a,R_a,t2_1_ab,t2_1_ab,optimize=True)
#######G^100#### block- ia
rdm1_a[:nocc_a,nocc_a:] = -np.einsum('n,ani->ia', R_a,L_aaa_u,optimize=True)
rdm1_a[:nocc_a,nocc_a:] += np.einsum('n,ain->ia', R_b,L_aba,optimize=True)
rdm1_b[:nocc_b,nocc_b:] = -np.einsum('n,ani->ia', R_b,L_bbb_u,optimize=True)
rdm1_b[:nocc_b,nocc_b:] += np.einsum('n,ain->ia', R_a,L_bab,optimize=True)
rdm1_a[:nocc_a,nocc_a:] -= np.einsum('g,cgh,ihac->ia', L_a,R_aaa_u,t2_1_a,optimize=True)
rdm1_a[:nocc_a,nocc_a:] += np.einsum('g,chg,ihac->ia', L_a,R_bab,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,nocc_a:] += np.einsum('g,chg,ihac->ia', L_b,R_aba,t2_1_a,optimize=True)
rdm1_a[:nocc_a,nocc_a:] -= np.einsum('g,cgh,ihac->ia', L_b,R_bbb_u,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,nocc_a:] += 0.5*np.einsum('i,cgh,ghac->ia', L_a,R_aaa_u,t2_1_a,optimize=True)
rdm1_a[:nocc_a,nocc_a:] -= np.einsum('i,chg,ghac->ia', L_a,R_bab,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,nocc_b:] -= np.einsum('g,cgh,ihac->ia', L_b,R_bbb_u,t2_1_b,optimize=True)
rdm1_b[:nocc_b,nocc_b:] += np.einsum('g,chg,hica->ia', L_b,R_aba,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,nocc_b:] += np.einsum('g,chg,hica->ia', L_a,R_bab,t2_1_b,optimize=True)
rdm1_b[:nocc_b,nocc_b:] -= np.einsum('g,cgh,hica->ia', L_a,R_aaa_u,t2_1_ab,optimize=True)
rdm1_b[:nocc_b,nocc_b:] += 0.5*np.einsum('i,cgh,ghac->ia', L_b,R_bbb_u,t2_1_b,optimize=True)
rdm1_b[:nocc_b,nocc_b:] -= np.einsum('i,chg,hgca->ia', L_b,R_aba,t2_1_ab,optimize=True)
rdm1_a[:nocc_a,nocc_a:] += np.einsum('g,g,ia->ia', L_a,R_a,t1_2_a,optimize=True)
rdm1_a[:nocc_a,nocc_a:] += np.einsum('g,g,ia->ia', L_b,R_b,t1_2_a,optimize=True)
rdm1_a[:nocc_a,nocc_a:] -= np.einsum('g,i,ga->ia', R_a,L_a,t1_2_a,optimize=True)
rdm1_b[:nocc_b,nocc_b:] += np.einsum('g,g,ia->ia', L_b,R_b,t1_2_b,optimize=True)
rdm1_b[:nocc_b,nocc_b:] += np.einsum('g,g,ia->ia', L_a,R_a,t1_2_b,optimize=True)
rdm1_b[:nocc_b,nocc_b:] -= np.einsum('g,i,ga->ia', R_b,L_b,t1_2_b,optimize=True)
############ block- ai
rdm1_a[nocc_a:,:nocc_a] = rdm1_a[:nocc_a,nocc_a:].T
rdm1_b[nocc_b:,:nocc_b] = rdm1_b[:nocc_b,nocc_b:].T
return (rdm1_a, rdm1_b)
[docs]
class UADCIP(uadc.UADC):
'''unrestricted ADC for IP energies and spectroscopic amplitudes
Attributes:
verbose : int
Print level. Default value equals to :class:`Mole.verbose`
max_memory : float or int
Allowed memory in MB. Default value equals to :class:`Mole.max_memory`
incore_complete : bool
Avoid all I/O. Default is False.
method : string
nth-order ADC method. Options are : ADC(2), ADC(2)-X, ADC(3). Default is ADC(2).
conv_tol : float
Convergence threshold for Davidson iterations. Default is 1e-12.
max_cycle : int
Number of Davidson iterations. Default is 50.
max_space : int
Space size to hold trial vectors for Davidson iterative diagonalization. Default is 12.
Kwargs:
nroots : int
Number of roots (eigenvalues) requested. Default value is 1.
>>> myadc = adc.UADC(mf).run()
>>> myadcip = adc.UADC(myadc).run()
Saved results
e_ip : float or list of floats
IP energy (eigenvalue). For nroots = 1, it is a single float
number. If nroots > 1, it is a list of floats for the lowest
nroots eigenvalues.
v_ip : array
Eigenvectors for each IP transition.
p_ip : float
Spectroscopic amplitudes for each IP transition.
'''
_keys = {
'tol_residual','conv_tol', 'e_corr', 'method',
'method_type', 'mo_coeff', 'mo_energy_b', 'max_memory',
't1', 'mo_energy_a', 'max_space', 't2', 'max_cycle',
'nocc_a', 'nocc_b', 'nvir_a', 'nvir_b', 'mo_coeff', 'mo_energy_a',
'mo_energy_b', 'nmo_a', 'nmo_b', 'mol', 'transform_integrals',
'with_df', 'spec_factor_print_tol', 'evec_print_tol',
'compute_properties', 'approx_trans_moments', 'E', 'U', 'P', 'X',
}
def __init__(self, adc):
self.mol = adc.mol
self.verbose = adc.verbose
self.stdout = adc.stdout
self.max_memory = adc.max_memory
self.max_space = adc.max_space
self.max_cycle = adc.max_cycle
self.conv_tol = adc.conv_tol
self.tol_residual = adc.tol_residual
self.t1 = adc.t1
self.t2 = adc.t2
self.imds = adc.imds
self.e_corr = adc.e_corr
self.method = adc.method
self.method_type = adc.method_type
self._scf = adc._scf
self._nocc = adc._nocc
self._nvir = adc._nvir
self._nmo = adc._nmo
self.nocc_a = adc._nocc[0]
self.nocc_b = adc._nocc[1]
self.nvir_a = adc._nvir[0]
self.nvir_b = adc._nvir[1]
self.mo_coeff = adc.mo_coeff
self.mo_energy_a = adc.mo_energy_a
self.mo_energy_b = adc.mo_energy_b
self.nmo_a = adc._nmo[0]
self.nmo_b = adc._nmo[1]
self.transform_integrals = adc.transform_integrals
self.with_df = adc.with_df
self.compute_properties = adc.compute_properties
self.approx_trans_moments = adc.approx_trans_moments
self.spec_factor_print_tol = adc.spec_factor_print_tol
self.evec_print_tol = adc.evec_print_tol
self.E = adc.E
self.U = adc.U
self.P = adc.P
self.X = adc.X
kernel = uadc.kernel
get_imds = get_imds
get_diag = get_diag
matvec = matvec
get_trans_moments = get_trans_moments
get_properties = get_properties
analyze_spec_factor = analyze_spec_factor
analyze_eigenvector = analyze_eigenvector
analyze = analyze
compute_dyson_mo = compute_dyson_mo
make_rdm1 = make_rdm1
[docs]
def get_init_guess(self, nroots=1, diag=None, ascending=True):
if diag is None :
diag = self.get_diag()
idx = None
if ascending:
idx = np.argsort(diag)
else:
idx = np.argsort(diag)[::-1]
guess = np.zeros((diag.shape[0], nroots))
min_shape = min(diag.shape[0], nroots)
guess[:min_shape,:min_shape] = np.identity(min_shape)
g = np.zeros((diag.shape[0], nroots))
g[idx] = guess.copy()
guess = []
for p in range(g.shape[1]):
guess.append(g[:,p])
return guess
[docs]
def gen_matvec(self, imds=None, eris=None):
if imds is None:
imds = self.get_imds(eris)
diag = self.get_diag(imds,eris)
matvec = self.matvec(imds, eris)
#matvec = lambda x: self.matvec()
return matvec, diag
