2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
5 * Copyright (c) 2001-2004, The GROMACS development team.
6 * Copyright (c) 2013,2014, by the GROMACS development team, led by
7 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
8 * and including many others, as listed in the AUTHORS file in the
9 * top-level source directory and at http://www.gromacs.org.
11 * GROMACS is free software; you can redistribute it and/or
12 * modify it under the terms of the GNU Lesser General Public License
13 * as published by the Free Software Foundation; either version 2.1
14 * of the License, or (at your option) any later version.
16 * GROMACS is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 * Lesser General Public License for more details.
21 * You should have received a copy of the GNU Lesser General Public
22 * License along with GROMACS; if not, see
23 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
24 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
26 * If you want to redistribute modifications to GROMACS, please
27 * consider that scientific software is very special. Version
28 * control is crucial - bugs must be traceable. We will be happy to
29 * consider code for inclusion in the official distribution, but
30 * derived work must not be called official GROMACS. Details are found
31 * in the README & COPYING files - if they are missing, get the
32 * official version at http://www.gromacs.org.
34 * To help us fund GROMACS development, we humbly ask that you cite
35 * the research papers on the package. Check out http://www.gromacs.org.
50 #include "chargegroup.h"
54 #include "gmx_fatal.h"
66 #include "gromacs/random/random.h"
71 #include "mtop_util.h"
73 #include "gromacs/gmxlib/conformation-utilities.h"
75 #include "gromacs/fileio/confio.h"
76 #include "gromacs/fileio/gmxfio.h"
77 #include "gromacs/fileio/trxio.h"
78 #include "gromacs/timing/wallcycle.h"
79 #include "gromacs/timing/walltime_accounting.h"
81 #ifdef GMX_SIMD_X86_SSE2_OR_HIGHER
82 #include "gromacs/simd/general_x86_sse2.h"
86 static void global_max(t_commrec *cr, int *n)
90 snew(sum, cr->nnodes);
92 gmx_sumi(cr->nnodes, sum, cr);
93 for (i = 0; i < cr->nnodes; i++)
101 static void realloc_bins(double **bin, int *nbin, int nbin_new)
105 if (nbin_new != *nbin)
107 srenew(*bin, nbin_new);
108 for (i = *nbin; i < nbin_new; i++)
116 double do_tpi(FILE *fplog, t_commrec *cr,
117 int nfile, const t_filenm fnm[],
118 const output_env_t oenv, gmx_bool bVerbose, gmx_bool gmx_unused bCompact,
119 int gmx_unused nstglobalcomm,
120 gmx_vsite_t gmx_unused *vsite, gmx_constr_t gmx_unused constr,
121 int gmx_unused stepout,
122 t_inputrec *inputrec,
123 gmx_mtop_t *top_global, t_fcdata *fcd,
126 t_nrnb *nrnb, gmx_wallcycle_t wcycle,
127 gmx_edsam_t gmx_unused ed,
129 int gmx_unused repl_ex_nst, int gmx_unused repl_ex_nex, int gmx_unused repl_ex_seed,
130 gmx_membed_t gmx_unused membed,
131 real gmx_unused cpt_period, real gmx_unused max_hours,
132 const char gmx_unused *deviceOptions,
133 unsigned long gmx_unused Flags,
134 gmx_walltime_accounting_t walltime_accounting)
136 const char *TPI = "Test Particle Insertion";
138 gmx_groups_t *groups;
139 gmx_enerdata_t *enerd;
141 real lambda, t, temp, beta, drmax, epot;
142 double embU, sum_embU, *sum_UgembU, V, V_all, VembU_all;
145 gmx_bool bDispCorr, bCharge, bRFExcl, bNotLastFrame, bStateChanged, bNS, bOurStep;
146 tensor force_vir, shake_vir, vir, pres;
147 int cg_tp, a_tp0, a_tp1, ngid, gid_tp, nener, e;
149 rvec mu_tot, x_init, dx, x_tp;
150 int nnodes, frame, nsteps, step;
154 char *ptr, *dump_pdb, **leg, str[STRLEN], str2[STRLEN];
155 double dbl, dump_ener;
157 int nat_cavity = 0, d;
158 real *mass_cavity = NULL, mass_tot;
160 double invbinw, *bin, refvolshift, logV, bUlogV;
161 real dvdl, prescorr, enercorr, dvdlcorr;
162 gmx_bool bEnergyOutOfBounds;
163 const char *tpid_leg[2] = {"direct", "reweighted"};
165 /* Since there is no upper limit to the insertion energies,
166 * we need to set an upper limit for the distribution output.
168 real bU_bin_limit = 50;
169 real bU_logV_bin_limit = bU_bin_limit + 10;
173 top = gmx_mtop_generate_local_top(top_global, inputrec);
175 groups = &top_global->groups;
177 bCavity = (inputrec->eI == eiTPIC);
180 ptr = getenv("GMX_TPIC_MASSES");
187 /* Read (multiple) masses from env var GMX_TPIC_MASSES,
188 * The center of mass of the last atoms is then used for TPIC.
191 while (sscanf(ptr, "%lf%n", &dbl, &i) > 0)
193 srenew(mass_cavity, nat_cavity+1);
194 mass_cavity[nat_cavity] = dbl;
195 fprintf(fplog, "mass[%d] = %f\n",
196 nat_cavity+1, mass_cavity[nat_cavity]);
202 gmx_fatal(FARGS, "Found %d masses in GMX_TPIC_MASSES", nat_cavity);
208 init_em(fplog,TPI,inputrec,&lambda,nrnb,mu_tot,
209 state->box,fr,mdatoms,top,cr,nfile,fnm,NULL,NULL);*/
210 /* We never need full pbc for TPI */
212 /* Determine the temperature for the Boltzmann weighting */
213 temp = inputrec->opts.ref_t[0];
216 for (i = 1; (i < inputrec->opts.ngtc); i++)
218 if (inputrec->opts.ref_t[i] != temp)
220 fprintf(fplog, "\nWARNING: The temperatures of the different temperature coupling groups are not identical\n\n");
221 fprintf(stderr, "\nWARNING: The temperatures of the different temperature coupling groups are not identical\n\n");
225 "\n The temperature for test particle insertion is %.3f K\n\n",
228 beta = 1.0/(BOLTZ*temp);
230 /* Number of insertions per frame */
231 nsteps = inputrec->nsteps;
233 /* Use the same neighborlist with more insertions points
234 * in a sphere of radius drmax around the initial point
236 /* This should be a proper mdp parameter */
237 drmax = inputrec->rtpi;
239 /* An environment variable can be set to dump all configurations
240 * to pdb with an insertion energy <= this value.
242 dump_pdb = getenv("GMX_TPI_DUMP");
246 sscanf(dump_pdb, "%lf", &dump_ener);
249 atoms2md(top_global, inputrec, 0, NULL, top_global->natoms, mdatoms);
250 update_mdatoms(mdatoms, inputrec->fepvals->init_lambda);
253 init_enerdata(groups->grps[egcENER].nr, inputrec->fepvals->n_lambda, enerd);
254 snew(f, top_global->natoms);
256 /* Print to log file */
257 walltime_accounting_start(walltime_accounting);
258 wallcycle_start(wcycle, ewcRUN);
259 print_start(fplog, cr, walltime_accounting, "Test Particle Insertion");
261 /* The last charge group is the group to be inserted */
262 cg_tp = top->cgs.nr - 1;
263 a_tp0 = top->cgs.index[cg_tp];
264 a_tp1 = top->cgs.index[cg_tp+1];
267 fprintf(debug, "TPI cg %d, atoms %d-%d\n", cg_tp, a_tp0, a_tp1);
269 if (a_tp1 - a_tp0 > 1 &&
270 (inputrec->rlist < inputrec->rcoulomb ||
271 inputrec->rlist < inputrec->rvdw))
273 gmx_fatal(FARGS, "Can not do TPI for multi-atom molecule with a twin-range cut-off");
275 snew(x_mol, a_tp1-a_tp0);
277 bDispCorr = (inputrec->eDispCorr != edispcNO);
279 for (i = a_tp0; i < a_tp1; i++)
281 /* Copy the coordinates of the molecule to be insterted */
282 copy_rvec(state->x[i], x_mol[i-a_tp0]);
283 /* Check if we need to print electrostatic energies */
284 bCharge |= (mdatoms->chargeA[i] != 0 ||
285 (mdatoms->chargeB && mdatoms->chargeB[i] != 0));
287 bRFExcl = (bCharge && EEL_RF(fr->eeltype) && fr->eeltype != eelRF_NEC);
289 calc_cgcm(fplog, cg_tp, cg_tp+1, &(top->cgs), state->x, fr->cg_cm);
292 if (norm(fr->cg_cm[cg_tp]) > 0.5*inputrec->rlist && fplog)
294 fprintf(fplog, "WARNING: Your TPI molecule is not centered at 0,0,0\n");
295 fprintf(stderr, "WARNING: Your TPI molecule is not centered at 0,0,0\n");
300 /* Center the molecule to be inserted at zero */
301 for (i = 0; i < a_tp1-a_tp0; i++)
303 rvec_dec(x_mol[i], fr->cg_cm[cg_tp]);
309 fprintf(fplog, "\nWill insert %d atoms %s partial charges\n",
310 a_tp1-a_tp0, bCharge ? "with" : "without");
312 fprintf(fplog, "\nWill insert %d times in each frame of %s\n",
313 nsteps, opt2fn("-rerun", nfile, fnm));
318 if (inputrec->nstlist > 1)
320 if (drmax == 0 && a_tp1-a_tp0 == 1)
322 gmx_fatal(FARGS, "Re-using the neighborlist %d times for insertions of a single atom in a sphere of radius %f does not make sense", inputrec->nstlist, drmax);
326 fprintf(fplog, "Will use the same neighborlist for %d insertions in a sphere of radius %f\n", inputrec->nstlist, drmax);
334 fprintf(fplog, "Will insert randomly in a sphere of radius %f around the center of the cavity\n", drmax);
338 ngid = groups->grps[egcENER].nr;
339 gid_tp = GET_CGINFO_GID(fr->cginfo[cg_tp]);
352 if (EEL_FULL(fr->eeltype))
357 snew(sum_UgembU, nener);
359 /* Initialize random generator */
360 tpi_rand = gmx_rng_init(inputrec->ld_seed);
364 fp_tpi = xvgropen(opt2fn("-tpi", nfile, fnm),
365 "TPI energies", "Time (ps)",
366 "(kJ mol\\S-1\\N) / (nm\\S3\\N)", oenv);
367 xvgr_subtitle(fp_tpi, "f. are averages over one frame", oenv);
370 sprintf(str, "-kT log(<Ve\\S-\\betaU\\N>/<V>)");
371 leg[e++] = strdup(str);
372 sprintf(str, "f. -kT log<e\\S-\\betaU\\N>");
373 leg[e++] = strdup(str);
374 sprintf(str, "f. <e\\S-\\betaU\\N>");
375 leg[e++] = strdup(str);
376 sprintf(str, "f. V");
377 leg[e++] = strdup(str);
378 sprintf(str, "f. <Ue\\S-\\betaU\\N>");
379 leg[e++] = strdup(str);
380 for (i = 0; i < ngid; i++)
382 sprintf(str, "f. <U\\sVdW %s\\Ne\\S-\\betaU\\N>",
383 *(groups->grpname[groups->grps[egcENER].nm_ind[i]]));
384 leg[e++] = strdup(str);
388 sprintf(str, "f. <U\\sdisp c\\Ne\\S-\\betaU\\N>");
389 leg[e++] = strdup(str);
393 for (i = 0; i < ngid; i++)
395 sprintf(str, "f. <U\\sCoul %s\\Ne\\S-\\betaU\\N>",
396 *(groups->grpname[groups->grps[egcENER].nm_ind[i]]));
397 leg[e++] = strdup(str);
401 sprintf(str, "f. <U\\sRF excl\\Ne\\S-\\betaU\\N>");
402 leg[e++] = strdup(str);
404 if (EEL_FULL(fr->eeltype))
406 sprintf(str, "f. <U\\sCoul recip\\Ne\\S-\\betaU\\N>");
407 leg[e++] = strdup(str);
410 xvgr_legend(fp_tpi, 4+nener, (const char**)leg, oenv);
411 for (i = 0; i < 4+nener; i++)
425 bNotLastFrame = read_first_frame(oenv, &status, opt2fn("-rerun", nfile, fnm),
426 &rerun_fr, TRX_NEED_X);
429 if (rerun_fr.natoms - (bCavity ? nat_cavity : 0) !=
430 mdatoms->nr - (a_tp1 - a_tp0))
432 gmx_fatal(FARGS, "Number of atoms in trajectory (%d)%s "
433 "is not equal the number in the run input file (%d) "
434 "minus the number of atoms to insert (%d)\n",
435 rerun_fr.natoms, bCavity ? " minus one" : "",
436 mdatoms->nr, a_tp1-a_tp0);
439 refvolshift = log(det(rerun_fr.box));
441 #ifdef GMX_SIMD_X86_SSE2_OR_HIGHER
442 /* Make sure we don't detect SSE overflow generated before this point */
443 gmx_mm_check_and_reset_overflow();
446 while (bNotLastFrame)
448 lambda = rerun_fr.lambda;
452 for (e = 0; e < nener; e++)
457 /* Copy the coordinates from the input trajectory */
458 for (i = 0; i < rerun_fr.natoms; i++)
460 copy_rvec(rerun_fr.x[i], state->x[i]);
462 copy_mat(rerun_fr.box, state->box);
467 bStateChanged = TRUE;
469 for (step = 0; step < nsteps; step++)
471 /* In parallel all nodes generate all random configurations.
472 * In that way the result is identical to a single cpu tpi run.
476 /* Random insertion in the whole volume */
477 bNS = (step % inputrec->nstlist == 0);
480 /* Generate a random position in the box */
481 x_init[XX] = gmx_rng_uniform_real(tpi_rand)*state->box[XX][XX];
482 x_init[YY] = gmx_rng_uniform_real(tpi_rand)*state->box[YY][YY];
483 x_init[ZZ] = gmx_rng_uniform_real(tpi_rand)*state->box[ZZ][ZZ];
485 if (inputrec->nstlist == 1)
487 copy_rvec(x_init, x_tp);
491 /* Generate coordinates within |dx|=drmax of x_init */
494 dx[XX] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
495 dx[YY] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
496 dx[ZZ] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
498 while (norm2(dx) > drmax*drmax);
499 rvec_add(x_init, dx, x_tp);
504 /* Random insertion around a cavity location
505 * given by the last coordinate of the trajectory.
511 /* Copy the location of the cavity */
512 copy_rvec(rerun_fr.x[rerun_fr.natoms-1], x_init);
516 /* Determine the center of mass of the last molecule */
519 for (i = 0; i < nat_cavity; i++)
521 for (d = 0; d < DIM; d++)
524 mass_cavity[i]*rerun_fr.x[rerun_fr.natoms-nat_cavity+i][d];
526 mass_tot += mass_cavity[i];
528 for (d = 0; d < DIM; d++)
530 x_init[d] /= mass_tot;
534 /* Generate coordinates within |dx|=drmax of x_init */
537 dx[XX] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
538 dx[YY] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
539 dx[ZZ] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
541 while (norm2(dx) > drmax*drmax);
542 rvec_add(x_init, dx, x_tp);
545 if (a_tp1 - a_tp0 == 1)
547 /* Insert a single atom, just copy the insertion location */
548 copy_rvec(x_tp, state->x[a_tp0]);
552 /* Copy the coordinates from the top file */
553 for (i = a_tp0; i < a_tp1; i++)
555 copy_rvec(x_mol[i-a_tp0], state->x[i]);
557 /* Rotate the molecule randomly */
558 rotate_conf(a_tp1-a_tp0, state->x+a_tp0, NULL,
559 2*M_PI*gmx_rng_uniform_real(tpi_rand),
560 2*M_PI*gmx_rng_uniform_real(tpi_rand),
561 2*M_PI*gmx_rng_uniform_real(tpi_rand));
562 /* Shift to the insertion location */
563 for (i = a_tp0; i < a_tp1; i++)
565 rvec_inc(state->x[i], x_tp);
569 /* Check if this insertion belongs to this node */
573 switch (inputrec->eI)
576 bOurStep = ((step / inputrec->nstlist) % nnodes == cr->nodeid);
579 bOurStep = (step % nnodes == cr->nodeid);
582 gmx_fatal(FARGS, "Unknown integrator %s", ei_names[inputrec->eI]);
587 /* Clear some matrix variables */
588 clear_mat(force_vir);
589 clear_mat(shake_vir);
593 /* Set the charge group center of mass of the test particle */
594 copy_rvec(x_init, fr->cg_cm[top->cgs.nr-1]);
596 /* Calc energy (no forces) on new positions.
597 * Since we only need the intermolecular energy
598 * and the RF exclusion terms of the inserted molecule occur
599 * within a single charge group we can pass NULL for the graph.
600 * This also avoids shifts that would move charge groups
603 * Some checks above ensure than we can not have
604 * twin-range interactions together with nstlist > 1,
605 * therefore we do not need to remember the LR energies.
607 /* Make do_force do a single node force calculation */
609 do_force(fplog, cr, inputrec,
610 step, nrnb, wcycle, top, &top_global->groups,
611 state->box, state->x, &state->hist,
612 f, force_vir, mdatoms, enerd, fcd,
614 NULL, fr, NULL, mu_tot, t, NULL, NULL, FALSE,
615 GMX_FORCE_NONBONDED | GMX_FORCE_ENERGY |
616 (bNS ? GMX_FORCE_DYNAMICBOX | GMX_FORCE_NS | GMX_FORCE_DO_LR : 0) |
617 (bStateChanged ? GMX_FORCE_STATECHANGED : 0));
619 bStateChanged = FALSE;
622 /* Calculate long range corrections to pressure and energy */
623 calc_dispcorr(fplog, inputrec, fr, step, top_global->natoms, state->box,
624 lambda, pres, vir, &prescorr, &enercorr, &dvdlcorr);
625 /* figure out how to rearrange the next 4 lines MRS 8/4/2009 */
626 enerd->term[F_DISPCORR] = enercorr;
627 enerd->term[F_EPOT] += enercorr;
628 enerd->term[F_PRES] += prescorr;
629 enerd->term[F_DVDL_VDW] += dvdlcorr;
631 epot = enerd->term[F_EPOT];
632 bEnergyOutOfBounds = FALSE;
633 #ifdef GMX_SIMD_X86_SSE2_OR_HIGHER
634 /* With SSE the energy can overflow, check for this */
635 if (gmx_mm_check_and_reset_overflow())
639 fprintf(debug, "Found an SSE overflow, assuming the energy is out of bounds\n");
641 bEnergyOutOfBounds = TRUE;
644 /* If the compiler doesn't optimize this check away
645 * we catch the NAN energies.
646 * The epot>GMX_REAL_MAX check catches inf values,
647 * which should nicely result in embU=0 through the exp below,
648 * but it does not hurt to check anyhow.
650 /* Non-bonded Interaction usually diverge at r=0.
651 * With tabulated interaction functions the first few entries
652 * should be capped in a consistent fashion between
653 * repulsion, dispersion and Coulomb to avoid accidental
654 * negative values in the total energy.
655 * The table generation code in tables.c does this.
656 * With user tbales the user should take care of this.
658 if (epot != epot || epot > GMX_REAL_MAX)
660 bEnergyOutOfBounds = TRUE;
662 if (bEnergyOutOfBounds)
666 fprintf(debug, "\n time %.3f, step %d: non-finite energy %f, using exp(-bU)=0\n", t, step, epot);
672 embU = exp(-beta*epot);
674 /* Determine the weighted energy contributions of each energy group */
676 sum_UgembU[e++] += epot*embU;
679 for (i = 0; i < ngid; i++)
682 (enerd->grpp.ener[egBHAMSR][GID(i, gid_tp, ngid)] +
683 enerd->grpp.ener[egBHAMLR][GID(i, gid_tp, ngid)])*embU;
688 for (i = 0; i < ngid; i++)
691 (enerd->grpp.ener[egLJSR][GID(i, gid_tp, ngid)] +
692 enerd->grpp.ener[egLJLR][GID(i, gid_tp, ngid)])*embU;
697 sum_UgembU[e++] += enerd->term[F_DISPCORR]*embU;
701 for (i = 0; i < ngid; i++)
704 (enerd->grpp.ener[egCOULSR][GID(i, gid_tp, ngid)] +
705 enerd->grpp.ener[egCOULLR][GID(i, gid_tp, ngid)])*embU;
709 sum_UgembU[e++] += enerd->term[F_RF_EXCL]*embU;
711 if (EEL_FULL(fr->eeltype))
713 sum_UgembU[e++] += enerd->term[F_COUL_RECIP]*embU;
718 if (embU == 0 || beta*epot > bU_bin_limit)
724 i = (int)((bU_logV_bin_limit
725 - (beta*epot - logV + refvolshift))*invbinw
733 realloc_bins(&bin, &nbin, i+10);
740 fprintf(debug, "TPI %7d %12.5e %12.5f %12.5f %12.5f\n",
741 step, epot, x_tp[XX], x_tp[YY], x_tp[ZZ]);
744 if (dump_pdb && epot <= dump_ener)
746 sprintf(str, "t%g_step%d.pdb", t, step);
747 sprintf(str2, "t: %f step %d ener: %f", t, step, epot);
748 write_sto_conf_mtop(str, str2, top_global, state->x, state->v,
749 inputrec->ePBC, state->box);
756 /* When running in parallel sum the energies over the processes */
757 gmx_sumd(1, &sum_embU, cr);
758 gmx_sumd(nener, sum_UgembU, cr);
763 VembU_all += V*sum_embU/nsteps;
767 if (bVerbose || frame%10 == 0 || frame < 10)
769 fprintf(stderr, "mu %10.3e <mu> %10.3e\n",
770 -log(sum_embU/nsteps)/beta, -log(VembU_all/V_all)/beta);
773 fprintf(fp_tpi, "%10.3f %12.5e %12.5e %12.5e %12.5e",
775 VembU_all == 0 ? 20/beta : -log(VembU_all/V_all)/beta,
776 sum_embU == 0 ? 20/beta : -log(sum_embU/nsteps)/beta,
778 for (e = 0; e < nener; e++)
780 fprintf(fp_tpi, " %12.5e", sum_UgembU[e]/nsteps);
782 fprintf(fp_tpi, "\n");
786 bNotLastFrame = read_next_frame(oenv, status, &rerun_fr);
787 } /* End of the loop */
788 walltime_accounting_end(walltime_accounting);
794 gmx_fio_fclose(fp_tpi);
799 fprintf(fplog, "\n");
800 fprintf(fplog, " <V> = %12.5e nm^3\n", V_all/frame);
801 fprintf(fplog, " <mu> = %12.5e kJ/mol\n", -log(VembU_all/V_all)/beta);
804 /* Write the Boltzmann factor histogram */
807 /* When running in parallel sum the bins over the processes */
810 realloc_bins(&bin, &nbin, i);
811 gmx_sumd(nbin, bin, cr);
815 fp_tpi = xvgropen(opt2fn("-tpid", nfile, fnm),
816 "TPI energy distribution",
817 "\\betaU - log(V/<V>)", "count", oenv);
818 sprintf(str, "number \\betaU > %g: %9.3e", bU_bin_limit, bin[0]);
819 xvgr_subtitle(fp_tpi, str, oenv);
820 xvgr_legend(fp_tpi, 2, (const char **)tpid_leg, oenv);
821 for (i = nbin-1; i > 0; i--)
823 bUlogV = -i/invbinw + bU_logV_bin_limit - refvolshift + log(V_all/frame);
824 fprintf(fp_tpi, "%6.2f %10d %12.5e\n",
827 bin[i]*exp(-bUlogV)*V_all/VembU_all);
829 gmx_fio_fclose(fp_tpi);
835 walltime_accounting_set_nsteps_done(walltime_accounting, frame*inputrec->nsteps);