src/mdlib/tpi.c

   1 /* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
   2  *
   3  *
   4  *                This source code is part of
   5  *
   6  *                 G   R   O   M   A   C   S
   7  *
   8  *          GROningen MAchine for Chemical Simulations
   9  *
  10  *                        VERSION 3.2.0
  11  * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
  12  * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
  13  * Copyright (c) 2001-2004, The GROMACS development team,
  14  * check out http://www.gromacs.org for more information.
  15
  16  * This program is free software; you can redistribute it and/or
  17  * modify it under the terms of the GNU General Public License
  18  * as published by the Free Software Foundation; either version 2
  19  * of the License, or (at your option) any later version.
  20  *
  21  * If you want to redistribute modifications, please consider that
  22  * scientific software is very special. Version control is crucial -
  23  * bugs must be traceable. We will be happy to consider code for
  24  * inclusion in the official distribution, but derived work must not
  25  * be called official GROMACS. Details are found in the README & COPYING
  26  * files - if they are missing, get the official version at www.gromacs.org.
  27  *
  28  * To help us fund GROMACS development, we humbly ask that you cite
  29  * the papers on the package - you can find them in the top README file.
  30  *
  31  * For more info, check our website at http://www.gromacs.org
  32  *
  33  * And Hey:
  34  * GROwing Monsters And Cloning Shrimps
  35  */
  36 #ifdef HAVE_CONFIG_H
  37 #include <config.h>
  38 #endif
  39
  40 #include <string.h>
  41 #include <time.h>
  42 #include <math.h>
  43 #include "sysstuff.h"
  44 #include "string2.h"
  45 #include "network.h"
  46 #include "confio.h"
  47 #include "copyrite.h"
  48 #include "smalloc.h"
  49 #include "nrnb.h"
  50 #include "main.h"
  51 #include "chargegroup.h"
  52 #include "force.h"
  53 #include "macros.h"
  54 #include "random.h"
  55 #include "names.h"
  56 #include "gmx_fatal.h"
  57 #include "txtdump.h"
  58 #include "typedefs.h"
  59 #include "update.h"
  60 #include "random.h"
  61 #include "constr.h"
  62 #include "vec.h"
  63 #include "statutil.h"
  64 #include "tgroup.h"
  65 #include "mdebin.h"
  66 #include "vsite.h"
  67 #include "force.h"
  68 #include "mdrun.h"
  69 #include "domdec.h"
  70 #include "partdec.h"
  71 #include "gmx_random.h"
  72 #include "physics.h"
  73 #include "xvgr.h"
  74 #include "mdatoms.h"
  75 #include "ns.h"
  76 #include "gmx_wallcycle.h"
  77 #include "mtop_util.h"
  78 #include "gmxfio.h"
  79 #include "pme.h"
  80 #include "gbutil.h"
  81
  82 #if ( defined(GMX_IA32_SSE) || defined(GMX_X86_64_SSE) || defined(GMX_X86_64_SSE2) )
  83 #if defined(GMX_DOUBLE)
  84 #include "gmx_sse2_double.h"
  85 #else
  86 #include "gmx_sse2_single.h"
  87 #endif
  88 #endif
  89
  90
  91 static void global_max(t_commrec *cr,int *n)
  92 {
  93   int *sum,i;
  94
  95   snew(sum,cr->nnodes);
  96   sum[cr->nodeid] = *n;
  97   gmx_sumi(cr->nnodes,sum,cr);
  98   for(i=0; i<cr->nnodes; i++)
  99     *n = max(*n,sum[i]);
 100
 101   sfree(sum);
 102 }
 103
 104 static void realloc_bins(double **bin,int *nbin,int nbin_new)
 105 {
 106   int i;
 107
 108   if (nbin_new != *nbin) {
 109     srenew(*bin,nbin_new);
 110     for(i=*nbin; i<nbin_new; i++)
 111       (*bin)[i] = 0;
 112     *nbin = nbin_new;
 113   }
 114 }
 115
 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 bCompact,
 119               int nstglobalcomm,
 120               gmx_vsite_t *vsite,gmx_constr_t constr,
 121               int stepout,
 122               t_inputrec *inputrec,
 123               gmx_mtop_t *top_global,t_fcdata *fcd,
 124               t_state *state,
 125               t_mdatoms *mdatoms,
 126               t_nrnb *nrnb,gmx_wallcycle_t wcycle,
 127               gmx_edsam_t ed,
 128               t_forcerec *fr,
 129               int repl_ex_nst,int repl_ex_seed,
 130               real cpt_period,real max_hours,
 131               const char *deviceOptions,
 132               unsigned long Flags,
 133               gmx_runtime_t *runtime)
 134 {
 135   const char *TPI="Test Particle Insertion";
 136   gmx_localtop_t *top;
 137   gmx_groups_t *groups;
 138   gmx_enerdata_t *enerd;
 139   rvec   *f;
 140   real   lambda,t,temp,beta,drmax,epot;
 141   double embU,sum_embU,*sum_UgembU,V,V_all,VembU_all;
 142   t_trxstatus   *status;
 143   t_trxframe rerun_fr;
 144   gmx_bool   bDispCorr,bCharge,bRFExcl,bNotLastFrame,bStateChanged,bNS,bOurStep;
 145   tensor force_vir,shake_vir,vir,pres;
 146   int    cg_tp,a_tp0,a_tp1,ngid,gid_tp,nener,e;
 147   rvec   *x_mol;
 148   rvec   mu_tot,x_init,dx,x_tp;
 149   int    nnodes,frame,nsteps,step;
 150   int    i,start,end;
 151   gmx_rng_t tpi_rand;
 152   FILE   *fp_tpi=NULL;
 153   char   *ptr,*dump_pdb,**leg,str[STRLEN],str2[STRLEN];
 154   double dbl,dump_ener;
 155   gmx_bool   bCavity;
 156   int    nat_cavity=0,d;
 157   real   *mass_cavity=NULL,mass_tot;
 158   int    nbin;
 159   double invbinw,*bin,refvolshift,logV,bUlogV;
 160   real dvdl,prescorr,enercorr,dvdlcorr;
 161   gmx_bool bEnergyOutOfBounds;
 162   const char *tpid_leg[2]={"direct","reweighted"};
 163
 164   /* Since there is no upper limit to the insertion energies,
 165    * we need to set an upper limit for the distribution output.
 166    */
 167   real bU_bin_limit      = 50;
 168   real bU_logV_bin_limit = bU_bin_limit + 10;
 169
 170   nnodes = cr->nnodes;
 171
 172   top = gmx_mtop_generate_local_top(top_global,inputrec);
 173
 174   groups = &top_global->groups;
 175
 176   bCavity = (inputrec->eI == eiTPIC);
 177   if (bCavity) {
 178     ptr = getenv("GMX_TPIC_MASSES");
 179     if (ptr == NULL) {
 180       nat_cavity = 1;
 181     } else {
 182       /* Read (multiple) masses from env var GMX_TPIC_MASSES,
 183        * The center of mass of the last atoms is then used for TPIC.
 184        */
 185       nat_cavity = 0;
 186       while (sscanf(ptr,"%lf%n",&dbl,&i) > 0) {
 187         srenew(mass_cavity,nat_cavity+1);
 188         mass_cavity[nat_cavity] = dbl;
 189         fprintf(fplog,"mass[%d] = %f\n",
 190                 nat_cavity+1,mass_cavity[nat_cavity]);
 191         nat_cavity++;
 192         ptr += i;
 193       }
 194       if (nat_cavity == 0)
 195         gmx_fatal(FARGS,"Found %d masses in GMX_TPIC_MASSES",nat_cavity);
 196     }
 197   }
 198
 199   /*
 200   init_em(fplog,TPI,inputrec,&lambda,nrnb,mu_tot,
 201   state->box,fr,mdatoms,top,cr,nfile,fnm,NULL,NULL);*/
 202   /* We never need full pbc for TPI */
 203   fr->ePBC = epbcXYZ;
 204   /* Determine the temperature for the Boltzmann weighting */
 205   temp = inputrec->opts.ref_t[0];
 206   if (fplog) {
 207     for(i=1; (i<inputrec->opts.ngtc); i++) {
 208       if (inputrec->opts.ref_t[i] != temp) {
 209         fprintf(fplog,"\nWARNING: The temperatures of the different temperature coupling groups are not identical\n\n");
 210         fprintf(stderr,"\nWARNING: The temperatures of the different temperature coupling groups are not identical\n\n");
 211       }
 212     }
 213     fprintf(fplog,
 214             "\n  The temperature for test particle insertion is %.3f K\n\n",
 215             temp);
 216   }
 217   beta = 1.0/(BOLTZ*temp);
 218
 219   /* Number of insertions per frame */
 220   nsteps = inputrec->nsteps;
 221
 222   /* Use the same neighborlist with more insertions points
 223    * in a sphere of radius drmax around the initial point
 224    */
 225   /* This should be a proper mdp parameter */
 226   drmax = inputrec->rtpi;
 227
 228   /* An environment variable can be set to dump all configurations
 229    * to pdb with an insertion energy <= this value.
 230    */
 231   dump_pdb = getenv("GMX_TPI_DUMP");
 232   dump_ener = 0;
 233   if (dump_pdb)
 234     sscanf(dump_pdb,"%lf",&dump_ener);
 235
 236   atoms2md(top_global,inputrec,0,NULL,0,top_global->natoms,mdatoms);
 237   update_mdatoms(mdatoms,inputrec->init_lambda);
 238
 239   snew(enerd,1);
 240   init_enerdata(groups->grps[egcENER].nr,inputrec->n_flambda,enerd);
 241   snew(f,top_global->natoms);
 242
 243   /* Print to log file  */
 244   runtime_start(runtime);
 245   print_date_and_time(fplog,cr->nodeid,
 246                       "Started Test Particle Insertion",runtime);
 247   wallcycle_start(wcycle,ewcRUN);
 248
 249   /* The last charge group is the group to be inserted */
 250   cg_tp = top->cgs.nr - 1;
 251   a_tp0 = top->cgs.index[cg_tp];
 252   a_tp1 = top->cgs.index[cg_tp+1];
 253   if (debug)
 254     fprintf(debug,"TPI cg %d, atoms %d-%d\n",cg_tp,a_tp0,a_tp1);
 255   if (a_tp1 - a_tp0 > 1 &&
 256       (inputrec->rlist < inputrec->rcoulomb ||
 257        inputrec->rlist < inputrec->rvdw))
 258     gmx_fatal(FARGS,"Can not do TPI for multi-atom molecule with a twin-range cut-off");
 259   snew(x_mol,a_tp1-a_tp0);
 260
 261   bDispCorr = (inputrec->eDispCorr != edispcNO);
 262   bCharge = FALSE;
 263   for(i=a_tp0; i<a_tp1; i++) {
 264     /* Copy the coordinates of the molecule to be insterted */
 265     copy_rvec(state->x[i],x_mol[i-a_tp0]);
 266     /* Check if we need to print electrostatic energies */
 267     bCharge |= (mdatoms->chargeA[i] != 0 ||
 268                 (mdatoms->chargeB && mdatoms->chargeB[i] != 0));
 269   }
 270   bRFExcl = (bCharge && EEL_RF(fr->eeltype) && fr->eeltype!=eelRF_NEC);
 271
 272   calc_cgcm(fplog,cg_tp,cg_tp+1,&(top->cgs),state->x,fr->cg_cm);
 273   if (bCavity) {
 274     if (norm(fr->cg_cm[cg_tp]) > 0.5*inputrec->rlist && fplog) {
 275       fprintf(fplog, "WARNING: Your TPI molecule is not centered at 0,0,0\n");
 276       fprintf(stderr,"WARNING: Your TPI molecule is not centered at 0,0,0\n");
 277     }
 278   } else {
 279     /* Center the molecule to be inserted at zero */
 280      for(i=0; i<a_tp1-a_tp0; i++)
 281       rvec_dec(x_mol[i],fr->cg_cm[cg_tp]);
 282   }
 283
 284   if (fplog) {
 285     fprintf(fplog,"\nWill insert %d atoms %s partial charges\n",
 286             a_tp1-a_tp0,bCharge ? "with" : "without");
 287
 288     fprintf(fplog,"\nWill insert %d times in each frame of %s\n",
 289             nsteps,opt2fn("-rerun",nfile,fnm));
 290   }
 291
 292     if (!bCavity)
 293     {
 294         if (inputrec->nstlist > 1)
 295         {
 296             if (drmax==0 && a_tp1-a_tp0==1)
 297             {
 298                 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);
 299             }
 300             if (fplog)
 301             {
 302                 fprintf(fplog,"Will use the same neighborlist for %d insertions in a sphere of radius %f\n",inputrec->nstlist,drmax);
 303             }
 304         }
 305     }
 306     else
 307     {
 308         if (fplog)
 309         {
 310             fprintf(fplog,"Will insert randomly in a sphere of radius %f around the center of the cavity\n",drmax);
 311         }
 312     }
 313
 314   ngid = groups->grps[egcENER].nr;
 315   gid_tp = GET_CGINFO_GID(fr->cginfo[cg_tp]);
 316   nener = 1 + ngid;
 317   if (bDispCorr)
 318     nener += 1;
 319   if (bCharge) {
 320     nener += ngid;
 321     if (bRFExcl)
 322       nener += 1;
 323     if (EEL_FULL(fr->eeltype))
 324       nener += 1;
 325   }
 326   snew(sum_UgembU,nener);
 327
 328   /* Initialize random generator */
 329   tpi_rand = gmx_rng_init(inputrec->ld_seed);
 330
 331   if (MASTER(cr)) {
 332     fp_tpi = xvgropen(opt2fn("-tpi",nfile,fnm),
 333                       "TPI energies","Time (ps)",
 334                       "(kJ mol\\S-1\\N) / (nm\\S3\\N)",oenv);
 335     xvgr_subtitle(fp_tpi,"f. are averages over one frame",oenv);
 336     snew(leg,4+nener);
 337     e = 0;
 338     sprintf(str,"-kT log(<Ve\\S-\\betaU\\N>/<V>)");
 339     leg[e++] = strdup(str);
 340     sprintf(str,"f. -kT log<e\\S-\\betaU\\N>");
 341     leg[e++] = strdup(str);
 342     sprintf(str,"f. <e\\S-\\betaU\\N>");
 343     leg[e++] = strdup(str);
 344     sprintf(str,"f. V");
 345     leg[e++] = strdup(str);
 346     sprintf(str,"f. <Ue\\S-\\betaU\\N>");
 347     leg[e++] = strdup(str);
 348     for(i=0; i<ngid; i++) {
 349       sprintf(str,"f. <U\\sVdW %s\\Ne\\S-\\betaU\\N>",
 350               *(groups->grpname[groups->grps[egcENER].nm_ind[i]]));
 351       leg[e++] = strdup(str);
 352     }
 353     if (bDispCorr) {
 354       sprintf(str,"f. <U\\sdisp c\\Ne\\S-\\betaU\\N>");
 355       leg[e++] = strdup(str);
 356     }
 357     if (bCharge) {
 358       for(i=0; i<ngid; i++) {
 359         sprintf(str,"f. <U\\sCoul %s\\Ne\\S-\\betaU\\N>",
 360                 *(groups->grpname[groups->grps[egcENER].nm_ind[i]]));
 361         leg[e++] = strdup(str);
 362       }
 363       if (bRFExcl) {
 364         sprintf(str,"f. <U\\sRF excl\\Ne\\S-\\betaU\\N>");
 365         leg[e++] = strdup(str);
 366       }
 367       if (EEL_FULL(fr->eeltype)) {
 368         sprintf(str,"f. <U\\sCoul recip\\Ne\\S-\\betaU\\N>");
 369         leg[e++] = strdup(str);
 370       }
 371     }
 372     xvgr_legend(fp_tpi,4+nener,(const char**)leg,oenv);
 373     for(i=0; i<4+nener; i++)
 374       sfree(leg[i]);
 375     sfree(leg);
 376   }
 377   clear_rvec(x_init);
 378   V_all = 0;
 379   VembU_all = 0;
 380
 381   invbinw = 10;
 382   nbin = 10;
 383   snew(bin,nbin);
 384
 385   bNotLastFrame = read_first_frame(oenv,&status,opt2fn("-rerun",nfile,fnm),
 386                                    &rerun_fr,TRX_NEED_X);
 387   frame = 0;
 388
 389   if (rerun_fr.natoms - (bCavity ? nat_cavity : 0) !=
 390       mdatoms->nr - (a_tp1 - a_tp0))
 391     gmx_fatal(FARGS,"Number of atoms in trajectory (%d)%s "
 392               "is not equal the number in the run input file (%d) "
 393               "minus the number of atoms to insert (%d)\n",
 394               rerun_fr.natoms,bCavity ? " minus one" : "",
 395               mdatoms->nr,a_tp1-a_tp0);
 396
 397   refvolshift = log(det(rerun_fr.box));
 398
 399 #if ( defined(GMX_IA32_SSE) || defined(GMX_X86_64_SSE) || defined(GMX_X86_64_SSE2) )
 400     /* Make sure we don't detect SSE overflow generated before this point */
 401     gmx_mm_check_and_reset_overflow();
 402 #endif
 403
 404     while (bNotLastFrame)
 405     {
 406         lambda = rerun_fr.lambda;
 407         t = rerun_fr.time;
 408
 409         sum_embU = 0;
 410         for(e=0; e<nener; e++)
 411         {
 412             sum_UgembU[e] = 0;
 413         }
 414
 415         /* Copy the coordinates from the input trajectory */
 416         for(i=0; i<rerun_fr.natoms; i++)
 417         {
 418             copy_rvec(rerun_fr.x[i],state->x[i]);
 419         }
 420
 421         V = det(rerun_fr.box);
 422         logV = log(V);
 423
 424         bStateChanged = TRUE;
 425         bNS = TRUE;
 426         for(step=0; step<nsteps; step++)
 427         {
 428             /* In parallel all nodes generate all random configurations.
 429              * In that way the result is identical to a single cpu tpi run.
 430              */
 431             if (!bCavity)
 432             {
 433                 /* Random insertion in the whole volume */
 434                 bNS = (step % inputrec->nstlist == 0);
 435                 if (bNS)
 436                 {
 437                     /* Generate a random position in the box */
 438                     x_init[XX] = gmx_rng_uniform_real(tpi_rand)*state->box[XX][XX];
 439                     x_init[YY] = gmx_rng_uniform_real(tpi_rand)*state->box[YY][YY];
 440                     x_init[ZZ] = gmx_rng_uniform_real(tpi_rand)*state->box[ZZ][ZZ];
 441                 }
 442                 if (inputrec->nstlist == 1)
 443                 {
 444                     copy_rvec(x_init,x_tp);
 445                 }
 446                 else
 447                 {
 448                     /* Generate coordinates within |dx|=drmax of x_init */
 449                     do
 450                     {
 451                         dx[XX] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
 452                         dx[YY] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
 453                         dx[ZZ] = (2*gmx_rng_uniform_real(tpi_rand) - 1)*drmax;
 454                     }
 455                     while (norm2(dx) > drmax*drmax);
 456                     rvec_add(x_init,dx,x_tp);
 457                 }
 458             }
 459             else
 460             {
 461                 /* Random insertion around a cavity location
 462                  * given by the last coordinate of the trajectory.
 463                  */
 464                 if (step == 0)
 465                 {
 466                     if (nat_cavity == 1)
 467                     {
 468                         /* Copy the location of the cavity */
 469                         copy_rvec(rerun_fr.x[rerun_fr.natoms-1],x_init);
 470                     }
 471                     else
 472                     {
 473                         /* Determine the center of mass of the last molecule */
 474                         clear_rvec(x_init);
 475                         mass_tot = 0;
 476                         for(i=0; i<nat_cavity; i++)
 477                         {
 478                             for(d=0; d<DIM; d++)
 479                             {
 480                                 x_init[d] +=
 481                                     mass_cavity[i]*rerun_fr.x[rerun_fr.natoms-nat_cavity+i][d];
 482                             }
 483                             mass_tot += mass_cavity[i];
 484                         }
 485                         for(d=0; d<DIM; d++)
 486                         {
 487                             x_init[d] /= mass_tot;
 488                         }
 489                     }
 490                 }
 491                 /* Generate coordinates within |dx|=drmax of x_init */
 492                 do
 493                 {
 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;
 497                 }
 498                 while (norm2(dx) > drmax*drmax);
 499                 rvec_add(x_init,dx,x_tp);
 500             }
 501
 502             if (a_tp1 - a_tp0 == 1)
 503             {
 504                 /* Insert a single atom, just copy the insertion location */
 505         copy_rvec(x_tp,state->x[a_tp0]);
 506             }
 507             else
 508             {
 509                 /* Copy the coordinates from the top file */
 510                 for(i=a_tp0; i<a_tp1; i++)
 511                 {
 512                     copy_rvec(x_mol[i-a_tp0],state->x[i]);
 513                 }
 514                 /* Rotate the molecule randomly */
 515                 rotate_conf(a_tp1-a_tp0,state->x+a_tp0,NULL,
 516                             2*M_PI*gmx_rng_uniform_real(tpi_rand),
 517                             2*M_PI*gmx_rng_uniform_real(tpi_rand),
 518                             2*M_PI*gmx_rng_uniform_real(tpi_rand));
 519                 /* Shift to the insertion location */
 520                 for(i=a_tp0; i<a_tp1; i++)
 521                 {
 522                     rvec_inc(state->x[i],x_tp);
 523                 }
 524             }
 525
 526             /* Check if this insertion belongs to this node */
 527             bOurStep = TRUE;
 528             if (PAR(cr))
 529             {
 530                 switch (inputrec->eI)
 531                 {
 532                 case eiTPI:
 533                     bOurStep = ((step / inputrec->nstlist) % nnodes == cr->nodeid);
 534                     break;
 535                 case eiTPIC:
 536                     bOurStep = (step % nnodes == cr->nodeid);
 537                     break;
 538                 default:
 539                     gmx_fatal(FARGS,"Unknown integrator %s",ei_names[inputrec->eI]);
 540                 }
 541             }
 542             if (bOurStep)
 543             {
 544                 /* Clear some matrix variables  */
 545                 clear_mat(force_vir);
 546                 clear_mat(shake_vir);
 547                 clear_mat(vir);
 548                 clear_mat(pres);
 549
 550                 /* Set the charge group center of mass of the test particle */
 551                 copy_rvec(x_init,fr->cg_cm[top->cgs.nr-1]);
 552
 553                 /* Calc energy (no forces) on new positions.
 554                  * Since we only need the intermolecular energy
 555                  * and the RF exclusion terms of the inserted molecule occur
 556                  * within a single charge group we can pass NULL for the graph.
 557                  * This also avoids shifts that would move charge groups
 558                  * out of the box.
 559                  *
 560                  * Some checks above ensure than we can not have
 561                  * twin-range interactions together with nstlist > 1,
 562                  * therefore we do not need to remember the LR energies.
 563                  */
 564                 /* Make do_force do a single node force calculation */
 565                 cr->nnodes = 1;
 566                 do_force(fplog,cr,inputrec,
 567                          step,nrnb,wcycle,top,top_global,&top_global->groups,
 568                          rerun_fr.box,state->x,&state->hist,
 569                          f,force_vir,mdatoms,enerd,fcd,
 570                          lambda,NULL,fr,NULL,mu_tot,t,NULL,NULL,FALSE,
 571                          GMX_FORCE_NONBONDED |
 572                          (bNS ? GMX_FORCE_NS | GMX_FORCE_DOLR : 0) |
 573                          (bStateChanged ? GMX_FORCE_STATECHANGED : 0));
 574                 cr->nnodes = nnodes;
 575                 bStateChanged = FALSE;
 576                 bNS = FALSE;
 577
 578                 /* Calculate long range corrections to pressure and energy */
 579                 calc_dispcorr(fplog,inputrec,fr,step,top_global->natoms,rerun_fr.box,
 580                               lambda,pres,vir,&prescorr,&enercorr,&dvdlcorr);
 581                 /* figure out how to rearrange the next 4 lines MRS 8/4/2009 */
 582                 enerd->term[F_DISPCORR] = enercorr;
 583                 enerd->term[F_EPOT] += enercorr;
 584                 enerd->term[F_PRES] += prescorr;
 585                 enerd->term[F_DVDL] += dvdlcorr;
 586
 587                 epot = enerd->term[F_EPOT];
 588                 bEnergyOutOfBounds = FALSE;
 589 #if ( defined(GMX_IA32_SSE) || defined(GMX_X86_64_SSE) || defined(GMX_X86_64_SSE2) )
 590                 /* With SSE the energy can overflow, check for this */
 591                 if (gmx_mm_check_and_reset_overflow())
 592                 {
 593                     if (debug)
 594                     {
 595                         fprintf(debug,"Found an SSE overflow, assuming the energy is out of bounds\n");
 596                     }
 597                     bEnergyOutOfBounds = TRUE;
 598                 }
 599 #endif
 600                 /* If the compiler doesn't optimize this check away
 601                  * we catch the NAN energies.
 602                  * The epot>GMX_REAL_MAX check catches inf values,
 603                  * which should nicely result in embU=0 through the exp below,
 604                  * but it does not hurt to check anyhow.
 605                  */
 606                 /* Non-bonded Interaction usually diverge at r=0.
 607                  * With tabulated interaction functions the first few entries
 608                  * should be capped in a consistent fashion between
 609                  * repulsion, dispersion and Coulomb to avoid accidental
 610                  * negative values in the total energy.
 611                  * The table generation code in tables.c does this.
 612                  * With user tbales the user should take care of this.
 613                  */
 614                 if (epot != epot || epot > GMX_REAL_MAX)
 615                 {
 616                     bEnergyOutOfBounds = TRUE;
 617                 }
 618                 if (bEnergyOutOfBounds)
 619                 {
 620                     if (debug)
 621                     {
 622                         fprintf(debug,"\n  time %.3f, step %d: non-finite energy %f, using exp(-bU)=0\n",t,step,epot);
 623                     }
 624                     embU = 0;
 625                 }
 626                 else
 627                 {
 628                     embU = exp(-beta*epot);
 629                     sum_embU += embU;
 630                     /* Determine the weighted energy contributions of each energy group */
 631                     e = 0;
 632                     sum_UgembU[e++] += epot*embU;
 633                     if (fr->bBHAM)
 634                     {
 635                         for(i=0; i<ngid; i++)
 636                         {
 637                             sum_UgembU[e++] +=
 638                                 (enerd->grpp.ener[egBHAMSR][GID(i,gid_tp,ngid)] +
 639                                  enerd->grpp.ener[egBHAMLR][GID(i,gid_tp,ngid)])*embU;
 640                         }
 641                     }
 642                     else
 643                     {
 644                         for(i=0; i<ngid; i++)
 645                         {
 646                             sum_UgembU[e++] +=
 647                                 (enerd->grpp.ener[egLJSR][GID(i,gid_tp,ngid)] +
 648                                  enerd->grpp.ener[egLJLR][GID(i,gid_tp,ngid)])*embU;
 649                         }
 650                     }
 651                     if (bDispCorr)
 652                     {
 653                         sum_UgembU[e++] += enerd->term[F_DISPCORR]*embU;
 654                     }
 655                     if (bCharge)
 656                     {
 657                         for(i=0; i<ngid; i++)
 658                         {
 659                             sum_UgembU[e++] +=
 660                                 (enerd->grpp.ener[egCOULSR][GID(i,gid_tp,ngid)] +
 661                                  enerd->grpp.ener[egCOULLR][GID(i,gid_tp,ngid)])*embU;
 662                         }
 663                         if (bRFExcl)
 664                         {
 665                             sum_UgembU[e++] += enerd->term[F_RF_EXCL]*embU;
 666                         }
 667                         if (EEL_FULL(fr->eeltype))
 668                         {
 669                             sum_UgembU[e++] += enerd->term[F_COUL_RECIP]*embU;
 670                         }
 671                     }
 672                 }
 673
 674                 if (embU == 0 || beta*epot > bU_bin_limit)
 675                 {
 676                     bin[0]++;
 677                 }
 678                 else
 679                 {
 680                     i = (int)((bU_logV_bin_limit
 681                                - (beta*epot - logV + refvolshift))*invbinw
 682                               + 0.5);
 683                     if (i < 0)
 684                     {
 685                         i = 0;
 686                     }
 687                     if (i >= nbin)
 688                     {
 689                         realloc_bins(&bin,&nbin,i+10);
 690                     }
 691                     bin[i]++;
 692                 }
 693
 694                 if (debug)
 695                 {
 696                     fprintf(debug,"TPI %7d %12.5e %12.5f %12.5f %12.5f\n",
 697                             step,epot,x_tp[XX],x_tp[YY],x_tp[ZZ]);
 698                 }
 699
 700                 if (dump_pdb && epot <= dump_ener)
 701                 {
 702                     sprintf(str,"t%g_step%d.pdb",t,step);
 703                     sprintf(str2,"t: %f step %d ener: %f",t,step,epot);
 704                     write_sto_conf_mtop(str,str2,top_global,state->x,state->v,
 705                                         inputrec->ePBC,state->box);
 706                 }
 707             }
 708         }
 709
 710         if (PAR(cr))
 711         {
 712             /* When running in parallel sum the energies over the processes */
 713             gmx_sumd(1,    &sum_embU, cr);
 714             gmx_sumd(nener,sum_UgembU,cr);
 715         }
 716
 717         frame++;
 718         V_all += V;
 719         VembU_all += V*sum_embU/nsteps;
 720
 721         if (fp_tpi)
 722         {
 723             if (bVerbose || frame%10==0 || frame<10)
 724             {
 725                 fprintf(stderr,"mu %10.3e <mu> %10.3e\n",
 726                         -log(sum_embU/nsteps)/beta,-log(VembU_all/V_all)/beta);
 727             }
 728
 729             fprintf(fp_tpi,"%10.3f %12.5e %12.5e %12.5e %12.5e",
 730                     t,
 731                     VembU_all==0 ? 20/beta : -log(VembU_all/V_all)/beta,
 732                     sum_embU==0  ? 20/beta : -log(sum_embU/nsteps)/beta,
 733                     sum_embU/nsteps,V);
 734             for(e=0; e<nener; e++)
 735             {
 736                 fprintf(fp_tpi," %12.5e",sum_UgembU[e]/nsteps);
 737             }
 738             fprintf(fp_tpi,"\n");
 739             fflush(fp_tpi);
 740         }
 741
 742         bNotLastFrame = read_next_frame(oenv, status,&rerun_fr);
 743     } /* End of the loop  */
 744     runtime_end(runtime);
 745
 746     close_trj(status);
 747
 748     if (fp_tpi != NULL)
 749     {
 750         gmx_fio_fclose(fp_tpi);
 751     }
 752
 753     if (fplog != NULL)
 754     {
 755         fprintf(fplog,"\n");
 756         fprintf(fplog,"  <V>  = %12.5e nm^3\n",V_all/frame);
 757         fprintf(fplog,"  <mu> = %12.5e kJ/mol\n",-log(VembU_all/V_all)/beta);
 758     }
 759
 760     /* Write the Boltzmann factor histogram */
 761     if (PAR(cr))
 762     {
 763         /* When running in parallel sum the bins over the processes */
 764         i = nbin;
 765         global_max(cr,&i);
 766         realloc_bins(&bin,&nbin,i);
 767         gmx_sumd(nbin,bin,cr);
 768     }
 769     if (MASTER(cr))
 770     {
 771         fp_tpi = xvgropen(opt2fn("-tpid",nfile,fnm),
 772                           "TPI energy distribution",
 773                           "\\betaU - log(V/<V>)","count",oenv);
 774         sprintf(str,"number \\betaU > %g: %9.3e",bU_bin_limit,bin[0]);
 775         xvgr_subtitle(fp_tpi,str,oenv);
 776         xvgr_legend(fp_tpi,2,(const char **)tpid_leg,oenv);
 777         for(i=nbin-1; i>0; i--)
 778         {
 779             bUlogV = -i/invbinw + bU_logV_bin_limit - refvolshift + log(V_all/frame);
 780             fprintf(fp_tpi,"%6.2f %10d %12.5e\n",
 781                     bUlogV,
 782                     (int)(bin[i]+0.5),
 783                     bin[i]*exp(-bUlogV)*V_all/VembU_all);
 784         }
 785         gmx_fio_fclose(fp_tpi);
 786     }
 787     sfree(bin);
 788
 789     sfree(sum_UgembU);
 790
 791     runtime->nsteps_done = frame*inputrec->nsteps;
 792
 793     return 0;
 794 }