[0b990d] | 1 | //
|
---|
| 2 | // hsosv1e1.cc
|
---|
| 3 | // based on: csgrade12.cc
|
---|
| 4 | //
|
---|
| 5 | // Copyright (C) 1996 Limit Point Systems, Inc.
|
---|
| 6 | //
|
---|
| 7 | // Author: Ida Nielsen <ida@kemi.aau.dk>
|
---|
| 8 | // Maintainer: LPS
|
---|
| 9 | //
|
---|
| 10 | // This file is part of the SC Toolkit.
|
---|
| 11 | //
|
---|
| 12 | // The SC Toolkit is free software; you can redistribute it and/or modify
|
---|
| 13 | // it under the terms of the GNU Library General Public License as published by
|
---|
| 14 | // the Free Software Foundation; either version 2, or (at your option)
|
---|
| 15 | // any later version.
|
---|
| 16 | //
|
---|
| 17 | // The SC Toolkit is distributed in the hope that it will be useful,
|
---|
| 18 | // but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 19 | // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 20 | // GNU Library General Public License for more details.
|
---|
| 21 | //
|
---|
| 22 | // You should have received a copy of the GNU Library General Public License
|
---|
| 23 | // along with the SC Toolkit; see the file COPYING.LIB. If not, write to
|
---|
| 24 | // the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
| 25 | //
|
---|
| 26 | // The U.S. Government is granted a limited license as per AL 91-7.
|
---|
| 27 | //
|
---|
| 28 |
|
---|
| 29 | #ifdef __GNUC__
|
---|
| 30 | #pragma implementation
|
---|
| 31 | #endif
|
---|
| 32 |
|
---|
| 33 | #include <math.h>
|
---|
| 34 |
|
---|
| 35 | #include <util/misc/formio.h>
|
---|
| 36 | #include <chemistry/qc/mbpt/bzerofast.h>
|
---|
| 37 | #include <chemistry/qc/mbpt/hsosv1e1.h>
|
---|
| 38 |
|
---|
| 39 | using namespace sc;
|
---|
| 40 |
|
---|
| 41 | #define PRINT1Q 0
|
---|
| 42 |
|
---|
| 43 | HSOSV1Erep1Qtr::HSOSV1Erep1Qtr(int mythread_a, int nthread_a,
|
---|
| 44 | int me_a, int nproc_a,
|
---|
| 45 | const Ref<ThreadLock> &lock_a,
|
---|
| 46 | const Ref<GaussianBasisSet> &basis_a,
|
---|
| 47 | const Ref<TwoBodyInt> &tbint_a,
|
---|
| 48 | int ni_a, double **scf_vector_a,
|
---|
| 49 | double tol_a, int debug_a)
|
---|
| 50 | {
|
---|
| 51 | mythread = mythread_a;
|
---|
| 52 | nthread = nthread_a;
|
---|
| 53 | me = me_a;
|
---|
| 54 | nproc = nproc_a;
|
---|
| 55 | lock = lock_a;
|
---|
| 56 | basis = basis_a;
|
---|
| 57 | tbint = tbint_a;
|
---|
| 58 | ni = ni_a;
|
---|
| 59 | scf_vector = scf_vector_a;
|
---|
| 60 | tol = tol_a;
|
---|
| 61 | debug = debug_a;
|
---|
| 62 |
|
---|
| 63 | nbasis = basis->nbasis();
|
---|
| 64 | nfuncmax = basis->max_nfunction_in_shell();
|
---|
| 65 | nshell = basis->nshell();
|
---|
| 66 |
|
---|
| 67 | aoint_computed_ = 0;
|
---|
| 68 |
|
---|
| 69 | trans_int1 = new double[nfuncmax*nfuncmax*nbasis*ni];
|
---|
| 70 |
|
---|
| 71 | timer = new RegionTimer();
|
---|
| 72 | }
|
---|
| 73 |
|
---|
| 74 | HSOSV1Erep1Qtr::~HSOSV1Erep1Qtr()
|
---|
| 75 | {
|
---|
| 76 | delete[] trans_int1;
|
---|
| 77 | }
|
---|
| 78 |
|
---|
| 79 | void
|
---|
| 80 | HSOSV1Erep1Qtr::accum_buffer(double *buffer)
|
---|
| 81 | {
|
---|
| 82 | int n = nr*ns*nbasis*ni;
|
---|
| 83 | for (int i=0; i<n; i++) {
|
---|
| 84 | buffer[i] += trans_int1[i];
|
---|
| 85 | }
|
---|
| 86 | }
|
---|
| 87 |
|
---|
| 88 | void
|
---|
| 89 | HSOSV1Erep1Qtr::set_data(int R_a,int nr_a,int S_a,int ns_a,int ni_a,int ioffset_a)
|
---|
| 90 | {
|
---|
| 91 | R = R_a;
|
---|
| 92 | nr = nr_a;
|
---|
| 93 | S = S_a;
|
---|
| 94 | ns = ns_a;
|
---|
| 95 | ni = ni_a;
|
---|
| 96 | i_offset = ioffset_a;
|
---|
| 97 | }
|
---|
| 98 |
|
---|
| 99 | void
|
---|
| 100 | HSOSV1Erep1Qtr::run()
|
---|
| 101 | {
|
---|
| 102 | int i;
|
---|
| 103 | int P, Q;
|
---|
| 104 | int bf1,bf2,bf3,bf4;
|
---|
| 105 | int p,q;
|
---|
| 106 | double *c_pi,*c_qi;
|
---|
| 107 |
|
---|
| 108 | timer->enter("bzerofast trans_int1");
|
---|
| 109 | bzerofast(trans_int1,nfuncmax*nfuncmax*nbasis*ni);
|
---|
| 110 | timer->exit("bzerofast trans_int1");
|
---|
| 111 |
|
---|
| 112 | const double *intbuf = tbint->buffer();
|
---|
| 113 |
|
---|
| 114 | int shell_index = 0;
|
---|
| 115 | int thindex = 0;
|
---|
| 116 |
|
---|
| 117 | for (P = 0; P < basis->nshell(); P++) {
|
---|
| 118 | int np = basis->shell(P).nfunction();
|
---|
| 119 |
|
---|
| 120 | for (Q = 0; Q <= P; Q++) {
|
---|
| 121 | shell_index++;
|
---|
| 122 | if (shell_index%nproc != me) continue;
|
---|
| 123 | if (thindex++%nthread != mythread) continue;
|
---|
| 124 |
|
---|
| 125 | if (tbint->log2_shell_bound(P,Q,R,S) < tol) {
|
---|
| 126 | continue; /* skip ereps less than tol */
|
---|
| 127 | }
|
---|
| 128 |
|
---|
| 129 | aoint_computed_++;
|
---|
| 130 |
|
---|
| 131 | int nq = basis->shell(Q).nfunction();
|
---|
| 132 |
|
---|
| 133 | timer->enter("erep");
|
---|
| 134 | tbint->compute_shell(P,Q,R,S);
|
---|
| 135 | timer->exit("erep");
|
---|
| 136 |
|
---|
| 137 | timer->enter("1. quart. tr.");
|
---|
| 138 |
|
---|
| 139 | int index = 0;
|
---|
| 140 |
|
---|
| 141 | for (bf1 = 0; bf1 < np; bf1++) {
|
---|
| 142 | p = basis->shell_to_function(P) + bf1;
|
---|
| 143 |
|
---|
| 144 | for (bf2 = 0; bf2 < nq; bf2++) {
|
---|
| 145 | q = basis->shell_to_function(Q) + bf2;
|
---|
| 146 | if (q > p) {
|
---|
| 147 | /* if q > p: want to skip the loops over bf3-4 */
|
---|
| 148 | /* and larger bf2 values, so increment bf1 by 1 */
|
---|
| 149 | /* ("break") and adjust the value of index */
|
---|
| 150 | index = (bf1 + 1) * nq * nr * ns;
|
---|
| 151 | break;
|
---|
| 152 | }
|
---|
| 153 |
|
---|
| 154 | for (bf3 = 0; bf3 < nr; bf3++) {
|
---|
| 155 |
|
---|
| 156 | for (bf4 = 0; bf4 < ns; bf4++,index++) {
|
---|
| 157 | if (R==S && bf4>bf3) {
|
---|
| 158 | index = ((bf1*nq + bf2)*nr + (bf3+1))*ns;
|
---|
| 159 | break;
|
---|
| 160 | }
|
---|
| 161 |
|
---|
| 162 | if (fabs(intbuf[index])>1.0e-15) {
|
---|
| 163 | double pqrs = intbuf[index];
|
---|
| 164 |
|
---|
| 165 | double *iqrs = &trans_int1[((bf4*nr + bf3)*nbasis + q)*ni];
|
---|
| 166 | double *iprs = &trans_int1[((bf4*nr + bf3)*nbasis + p)*ni];
|
---|
| 167 |
|
---|
| 168 | if (p == q) pqrs *= 0.5;
|
---|
| 169 |
|
---|
| 170 | int col_index = i_offset;
|
---|
| 171 | c_pi = &scf_vector[p][col_index];
|
---|
| 172 | c_qi = &scf_vector[q][col_index];
|
---|
| 173 |
|
---|
| 174 | for (i=ni; i; i--) {
|
---|
| 175 | *iqrs++ += pqrs * *c_pi++;
|
---|
| 176 | *iprs++ += pqrs * *c_qi++;
|
---|
| 177 | }
|
---|
| 178 | }
|
---|
| 179 | } /* exit bf4 loop */
|
---|
| 180 | } /* exit bf3 loop */
|
---|
| 181 | } /* exit bf2 loop */
|
---|
| 182 | } /* exit bf1 loop */
|
---|
| 183 | timer->exit("1. quart. tr.");
|
---|
| 184 | } /* exit Q loop */
|
---|
| 185 | } /* exit P loop */
|
---|
| 186 | }
|
---|
| 187 |
|
---|
| 188 | ////////////////////////////////////////////////////////////////////////////
|
---|
| 189 |
|
---|
| 190 | // Local Variables:
|
---|
| 191 | // mode: c++
|
---|
| 192 | // c-file-style: "CLJ-CONDENSED"
|
---|
| 193 | // End:
|
---|