#include <algorithm>
#include "CompressPsf.h"
#include "strlib.h"
#include "Molecule.h"
#include "Parameters.h"
#include "SimParameters.h"
#include "InfoStream.h"
#include "UniqueSet.h"
#include "UniqueSetIter.h"

Classes
struct	BasicAtomInfo

struct	AtomSigInfo

struct	ExclSigInfo

class	HashString

class	HashReal

Functions
int	operator== (const AtomSigInfo &s1, const AtomSigInfo &s2)

int	operator== (const ExclSigInfo &s1, const ExclSigInfo &s2)

int	operator== (const BondValue &b1, const BondValue &b2)

int	operator== (const AngleValue &a1, const AngleValue &a2)

int	operator!= (const FourBodyConsts &f1, const FourBodyConsts &f2)

int	operator== (const DihedralValue &d1, const DihedralValue &d2)

int	operator== (const ImproperValue &d1, const ImproperValue &d2)

void	loadMolInfo ()

void	integrateAllAtomSigs ()

void	outputCompressedFile (FILE txtOfp, FILE binOfp)

void	getExtraBonds (StringList *file)

void	buildAtomData ()

void	buildBondData ()

void	buildAngleData ()

void	buildDihedralData ()

void	buildImproperData ()

void	buildCrosstermData ()

void	buildExclusionData ()

void	buildExclusions ()

void	build12Excls (UniqueSet< Exclusion > &, vector< int > *)

void	build13Excls (UniqueSet< Exclusion > &, vector< int > *)

void	build14Excls (UniqueSet< Exclusion > &, vector< int > *, int)

void	flipNum (char *elem, int elemSize, int numElems)

void	clearGlobalVectors ()

void	compress_molecule_info (Molecule mol, char psfFileName, Parameters param, SimParameters simParam, ConfigList *cfgList)

Variables
Molecule *	g_mol = NULL

Parameters *	g_param = NULL

SimParameters *	g_simParam = NULL

ConfigList *	g_cfgList = NULL

HashPool< HashString >	segNamePool

HashPool< HashString >	resNamePool

HashPool< HashString >	atomNamePool

HashPool< HashString >	atomTypePool

HashPool< HashReal >	chargePool

HashPool< HashReal >	massPool

HashPool< AtomSigInfo >	atomSigPool

BasicAtomInfo *	atomData

int *	eachAtomClusterID = NULL

vector< int >	eachClusterSize

vector< int >	eachClusterID

int	g_numClusters = 0

HashPool< TupleSignature >	sigsOfBonds

HashPool< TupleSignature >	sigsOfAngles

HashPool< TupleSignature >	sigsOfDihedrals

HashPool< TupleSignature >	sigsOfImpropers

HashPool< TupleSignature >	sigsOfCrossterms

AtomSigInfo *	eachAtomSigs

HashPool< ExclSigInfo >	sigsOfExclusions

ExclSigInfo *	eachAtomExclSigs

vector< Bond >	extraBonds

vector< Angle >	extraAngles

vector< Dihedral >	extraDihedrals

vector< Improper >	extraImpropers

vector< BondValue >	extraBondParams

vector< AngleValue >	extraAngleParams

vector< DihedralValue >	extraDihedralParams

vector< ImproperValue >	extraImproperParams

Function Documentation

◆ build12Excls()

void build12Excls	(	UniqueSet< Exclusion > &	allExcls,
		vector< int > *	eachAtomNeighbors
	)

Definition at line 1314 of file CompressPsf.C.

References UniqueSet< Elem >::add(), g_mol, and Molecule::numAtoms.

Referenced by buildExclusions().

 {
     for(int atom1=0; atom1<g_mol->numAtoms; atom1++)
     {
         vector<int> *atom1List = &eachAtomNeighbors[atom1];
         for(int j=0; j<atom1List->size(); j++)
         {
             int atom2 = atom1List->at(j);
             if(atom1<atom2)
                 allExcls.add(Exclusion(atom1, atom2));
             else
                 allExcls.add(Exclusion(atom2, atom1));
         }
     }
 }

◆ build13Excls()

void build13Excls	(	UniqueSet< Exclusion > &	allExcls,
		vector< int > *	eachAtomNeighbors
	)

Definition at line 1330 of file CompressPsf.C.

References UniqueSet< Elem >::add(), g_mol, and Molecule::numAtoms.

Referenced by buildExclusions().

 {
     for(int atom1=0; atom1<g_mol->numAtoms; atom1++)
     {
         vector<int> *atom1List = &eachAtomNeighbors[atom1];
         for(int j=0; j<atom1List->size(); j++)
         {
             int atom2 = atom1List->at(j);
             vector<int> *atom2List = &eachAtomNeighbors[atom2];
             for(int k=0; k<atom2List->size(); k++)
             {
                 int atom3 = atom2List->at(k);
                 //atom1-atom2, so atom2List contains atom1 which should not be considered
                 if(atom3 == atom1)
                     continue;
                 if(atom1<atom3)
                     allExcls.add(Exclusion(atom1, atom3));
                 else
                     allExcls.add(Exclusion(atom3, atom1));
             }
         }
     }
 }

◆ build14Excls()

void build14Excls	(	UniqueSet< Exclusion > &	allExcls,
		vector< int > *	eachAtomNeighbors,
		int	modified
	)

Definition at line 1354 of file CompressPsf.C.

References UniqueSet< Elem >::add(), g_mol, and Molecule::numAtoms.

Referenced by buildExclusions().

 {
     for(int atom1=0; atom1<g_mol->numAtoms; atom1++)
     {
         vector<int> *atom1List = &eachAtomNeighbors[atom1];
         for(int j=0; j<atom1List->size(); j++)
         {
             int atom2 = atom1List->at(j);
             vector<int> *atom2List = &eachAtomNeighbors[atom2];
             for(int k=0; k<atom2List->size(); k++)
             {
                 int atom3 = atom2List->at(k);
                 //atom1-atom2, so atom2List contains atom1 which should not be considered
                 if(atom3 == atom1)
                     continue;
                 vector<int> *atom3List = &eachAtomNeighbors[atom3];
                 for(int l=0; l<atom3List->size(); l++)
                 {
                     int atom4 = atom3List->at(l);
                     //atom1-atom2, so atom2List contains atom1 which should not be considered
                     if(atom4 == atom2 || atom4 == atom1)
                         continue;
                     if(atom1<atom4)
                         allExcls.add(Exclusion(atom1, atom4, modified));
                     else
                         allExcls.add(Exclusion(atom4, atom1, modified));
                 }
             }
         }
     }
 }

◆ buildAngleData()

void buildAngleData ( )

Definition at line 1081 of file CompressPsf.C.

References ANGLE, angle::angle_type, AtomSigInfo::angleSigIndices, angle::atom1, angle::atom2, angle::atom3, eachAtomSigs, g_mol, Molecule::getAllAngles(), Molecule::numAngles, TupleSignature::setOffsets(), and sigsOfAngles.

Referenced by loadMolInfo().

 {
 #ifndef MEM_OPT_VERSION
     Angle *angles = g_mol->getAllAngles();
     //create angles' tupleSignature
     for(int i=0; i<g_mol->numAngles; i++)
     {
         Angle *tuple = angles+i;
         TupleSignature oneSig(2,ANGLE,tuple->angle_type);
         int offset[2];
         offset[0] = tuple->atom2 - tuple->atom1;
         offset[1] = tuple->atom3 - tuple->atom1;
         oneSig.setOffsets(offset);
 
         int poolIndex = sigsOfAngles.lookupCstPool(oneSig);
         if(poolIndex == -1)
         {
             sigsOfAngles.push_back(oneSig);
             poolIndex = (SigIndex)sigsOfAngles.size()-1;
         }
         eachAtomSigs[tuple->atom1].angleSigIndices.push_back(poolIndex);
     }
     delete [] angles;
 #endif
 }

◆ buildAtomData()

void buildAtomData ( )

Definition at line 764 of file CompressPsf.C.

References atomNamePool, atomTypePool, chargePool, g_mol, Molecule::getAtomNames(), Molecule::getAtoms(), Molecule::getAtomSegResInfo(), massPool, Molecule::numAtoms, seg_resid::resid, resNamePool, and segNamePool.

Referenced by loadMolInfo().

 {
 #ifndef MEM_OPT_VERSION
     int numAtoms = g_mol->numAtoms;
 
     //1. parse atom data to build constant pool (atom name, mass, charge etc.)
     atomData = new BasicAtomInfo[numAtoms];
     Atom *atoms = g_mol->getAtoms();
     AtomNameInfo *atomNames = g_mol->getAtomNames();
     AtomSegResInfo *atomSegResids = g_mol->getAtomSegResInfo();
 
     for(int atomID=0; atomID < numAtoms; atomID++)
     {
         //building constant pool
         int poolIndex;
         HashString fieldName;
         fieldName.assign(atomSegResids[atomID].segname);
         poolIndex = segNamePool.lookupCstPool(fieldName);
         if(poolIndex==-1)
         {
             segNamePool.push_back(fieldName);
             poolIndex = segNamePool.size()-1;
         }
         atomData[atomID].segNameIdx = poolIndex;
         
         atomData[atomID].resID = atomSegResids[atomID].resid;
 
         fieldName.assign(atomNames[atomID].resname);
         poolIndex = resNamePool.lookupCstPool(fieldName);
         if(poolIndex==-1)
         {
             resNamePool.push_back(fieldName);
             poolIndex = resNamePool.size()-1;
         }
         atomData[atomID].resNameIdx = poolIndex;
 
         fieldName.assign(atomNames[atomID].atomname);
         poolIndex = atomNamePool.lookupCstPool(fieldName);
         if(poolIndex==-1)
         {
             atomNamePool.push_back(fieldName);
             poolIndex = atomNamePool.size()-1;
         }
         atomData[atomID].atomNameIdx = poolIndex;
 
         fieldName.assign(atomNames[atomID].atomtype);
         poolIndex = atomTypePool.lookupCstPool(fieldName);
         if(poolIndex==-1)
         {
             atomTypePool.push_back(fieldName);
             poolIndex = atomTypePool.size()-1;
         }
         atomData[atomID].atomTypeIdx = poolIndex;
 
         poolIndex = chargePool.lookupCstPool(atoms[atomID].charge);
         if(poolIndex==-1)
         {
             chargePool.push_back(atoms[atomID].charge);
             poolIndex = chargePool.size()-1;
         }
         atomData[atomID].chargeIdx = poolIndex;
 
         poolIndex = massPool.lookupCstPool(atoms[atomID].mass);
         if(poolIndex==-1)
         {
             massPool.push_back(atoms[atomID].mass);
             poolIndex = massPool.size()-1;
         }
         atomData[atomID].massIdx = poolIndex;
     }
 
     //Free those space to reduce transient memory usage
     //delete [] atoms; (deleted until per-atom info is output)
     delete [] atomNames;
     delete [] atomSegResids;
 #endif
 }

◆ buildBondData()

void buildBondData ( )

Definition at line 843 of file CompressPsf.C.

References bond::atom1, bond::atom2, BOND, bond::bond_type, AtomSigInfo::bondSigIndices, eachAtomClusterID, eachAtomSigs, eachClusterID, eachClusterSize, g_mol, g_numClusters, Molecule::getAllBonds(), TupleSignature::isReal, lookupCstPool(), NAMD_die(), Molecule::numAtoms, Molecule::numBonds, Molecule::numRealBonds, TupleSignature::offset, and sigsOfBonds.

Referenced by loadMolInfo().

 {
 #ifndef MEM_OPT_VERSION
     Bond *bonds = g_mol->getAllBonds();    
 
     //then creating bond's tupleSignature
     for(int i=0; i<g_mol->numBonds; i++)
     {
         Bond *b = bonds+i;
         TupleSignature oneSig(1,BOND,b->bond_type);
         oneSig.offset[0] = b->atom2 - b->atom1;
         oneSig.isReal = (i<g_mol->numRealBonds);
 
         int poolIndex = sigsOfBonds.lookupCstPool(oneSig);
         int newSig=0;
         if(poolIndex == -1)
         {
             sigsOfBonds.push_back(oneSig);
             poolIndex = (SigIndex)sigsOfBonds.size()-1;
             newSig=1;
         }
 
       if(!newSig)
         {//check duplicate bonds in the form of (a, b) && (a, b);
           int dupIdx = lookupCstPool(eachAtomSigs[b->atom1].bondSigIndices, (SigIndex)poolIndex);
             if(dupIdx!=-1)
             {
                 char err_msg[128];
                 sprintf(err_msg, "Duplicate bond %d-%d!", b->atom1+1, b->atom2+1);
                 NAMD_die(err_msg);
             }
         }
         eachAtomSigs[b->atom1].bondSigIndices.push_back(poolIndex);
     }
 
     //check duplicate bonds in the form of (a, b) && (b, a)
     for(int i=0; i<g_mol->numBonds; i++)
     {
         Bond *b=bonds+i;
         int atom2 = b->atom2;
         int thisOffset = atom2 - b->atom1;
         for(int j=0; j<eachAtomSigs[atom2].bondSigIndices.size(); j++)
         {
             SigIndex atom2BondId = eachAtomSigs[atom2].bondSigIndices[j];
             TupleSignature *secSig = &(sigsOfBonds[atom2BondId]);
             if(thisOffset== -(secSig->offset[0]))
             {
                 char err_msg[128];
                 sprintf(err_msg, "Duplicate bond %d-%d because two atoms are just reversed!", b->atom1+1, atom2+1);
                 NAMD_die(err_msg);
             }
         }
     }      
 
     //building clusters for this simulation system in two steps
     //1. create a list for each atom where each atom in the list is bonded with that atom
     vector<int> *atomListOfBonded = new vector<int>[g_mol->numAtoms];
 
     for(int i=0; i<g_mol->numRealBonds; i++)
     {
         Bond *b=bonds+i;
         int atom1 = b->atom1;
         int atom2 = b->atom2;
         atomListOfBonded[atom1].push_back(atom2);
         atomListOfBonded[atom2].push_back(atom1);
     }
 
     delete [] bonds;
 
     //2. using breadth-first-search to build the clusters. Here, we avoid recursive call
     // because the depth of calls may be of thousands which will blow up the stack, and
     //recursive call is slower than the stack-based BFS.
     //Considering such structure
     //1->1245; 7->1243; 1243->1245
     eachAtomClusterID = new int[g_mol->numAtoms];
     for(int i=0; i<g_mol->numAtoms; i++)
         eachAtomClusterID[i] = -1;
 
     //It is guaranteed that the clusters found in this way use the
     //smallest atom id of this cluster because each atom is at least
     //connected to one of the atoms in its cluster (by atomListOfBonded
     //constructed above).
     //--Chao Mei
     for(int i=0; i<g_mol->numAtoms; i++)
     {
         int curClusterID=eachAtomClusterID[i];
         //if the atom's cluster id is not -1, the atom has been visited
         if(curClusterID!=-1) continue;
 
         curClusterID=i;
         deque<int> toVisitAtoms;
         eachAtomClusterID[i] = curClusterID;
         toVisitAtoms.push_back(i);
         while(!toVisitAtoms.empty())
         {
             int visAtomID = toVisitAtoms.front();
             toVisitAtoms.pop_front();
             for(int j=0; j<atomListOfBonded[visAtomID].size(); j++)
             {
                 int otherAtom = atomListOfBonded[visAtomID][j];
                 if(eachAtomClusterID[otherAtom]!=curClusterID){
                     eachAtomClusterID[otherAtom]=curClusterID;
                     toVisitAtoms.push_back(otherAtom);
                 }
             }
         }
     }
 
 #if 0
     //Now the clusterID of each atom should be usually in the non-decreasing
     //order. In other words, the atom ids of a cluster are generally contiguous.
     //If this is the case, the temporary memory usage of output IO during
     //the simulation can be dramatically reduced. So g_isClusterContiguous
     //is used to differentiate the two cases: 
 
     //1. if the cluster id of atoms is monotonically increasing 
     //(g_isClusterContiguous=1), the size of the cluster can be used as this
     //this cluster's signature (represented by "eachAtomClusterID").
     //The atom who is the first atom (in terms of atom id) in this cluster will 
     //store the cluster size as its signature. The remaining atoms in this 
     //cluster store -1.
     
     //2. if the cluster id of atoms is not monotonically increasing, that is,
     //the atom ids of a cluster are not contiguous (g_isClusterContiguous=0).
     //Then we have to still record each atom's cluster id in variable 
     //"eachAtomClusterID", and we have to change each unique cluster id (valued
     //at the scale of numAtoms) into indexes at the scale of numClusters. For
     //example, the cluster ids of atoms up to this point may be (0...0, 24...24,
     //789...789,...), after re-scaling, cluster ids should be (0...0,1...1,
     //2...2,....) for atoms.
     //We made an ASSUMPTION here: the cluster ids are in the non-decreasing
     //order, and when a cluster discontiguity happens, the cluster id has
     //appeared before! Such ASSUMPTION is to make implementation easy.
     
     int curClusterID;
     int prevClusterID = eachAtomClusterID[0];
     int curClusterSize = 1;
     //step1: segment all atoms according to each cluster
     for(int i=1; i<g_mol->numAtoms; i++){
         curClusterID = eachAtomClusterID[i];
         if(curClusterID == prevClusterID){
                 curClusterSize++;
         }else{
                 eachClusterSize.push_back(curClusterSize);
                 eachClusterID.push_back(prevClusterID);
                 curClusterSize=1;
         }
         prevClusterID = curClusterID;   
     }
     //record the info of the last cluster segment
     eachClusterSize.push_back(curClusterSize);
     eachClusterID.push_back(prevClusterID);
     
     //step2: detect contiguity of atoms in each cluster and re-scale
     //cluster id.
     g_isClusterContiguous = 1;    
     int *newClusterIDs = new int[eachClusterID.size()];
     memset(newClusterIDs, 0, sizeof(int)*eachClusterID.size());
     prevClusterID = eachClusterID[0];
     int newCId = 0;    
     newClusterIDs[0] = newCId;
     for(int seg=1; seg<eachClusterID.size(); seg++){
         curClusterID = eachClusterID[seg];
         if(curClusterID > prevClusterID){
                 newClusterIDs[seg] = ++newCId;
                 prevClusterID = curClusterID;
         }else{
                 //non-contiguity happens                
                 g_isClusterContiguous = 0;
                 
                 //we ASSUME this id appears before
                 //binary search in eachAtomClusterID[0...seg-1).
                 int jl=0, jh=seg-2;
                 int isFound = 0;
                 while(jh>=jl){
                         int mid = (jl+jh)/2;
                         if(curClusterID > eachClusterID[mid])
                                 jl = mid+1;
                         else if(curClusterID < eachClusterID[mid])
                                 jh = mid-1;
                         else{
                                 newClusterIDs[seg] = newClusterIDs[mid];                                
                                 isFound = 1;
                                 break;
                         }                       
                 }
                 if(!isFound){
                         //ASSUMPTION is wrong and abort
                         char errmsg[300];
                         sprintf(errmsg, "Assumption about building cluster is broken in file %s at line %d\n", __FILE__, __LINE__);
                         NAMD_die(errmsg);
                 }                               
         }       
     }
     
     //step 3: modify eachAtomClusterID according to g_isClusterContiguous
     //newCId is the id of the last cluster, as id starts from 0, so the
     //total number clusters should be newCId+1
     g_numClusters = newCId+1;
     if(g_isClusterContiguous){
         int aid=0;
         for(int seg=0; seg<eachClusterSize.size(); seg++)
         {
             int curSize = eachClusterSize[seg];
             eachAtomClusterID[aid] = curSize;
             for(int i=aid+1; i<aid+curSize; i++)
                 eachAtomClusterID[i] = -1;
             aid += curSize;
         }       
     }else{
         int aid=0;
         for(int seg=0; seg<eachClusterSize.size(); seg++)
         {
             int curSize = eachClusterSize[seg];            
             for(int i=aid; i<aid+curSize; i++)
                 eachAtomClusterID[i] = newClusterIDs[seg];
             aid += curSize;
         }
     }
     free(newClusterIDs);
     eachClusterSize.clear();
     eachClusterID.clear();
 #endif 
                 
 /*
     //check whether cluster is built correctly
     printf("num clusters: %d\n", g_numClusters);
     FILE *checkFile = fopen("cluster.opt", "w");
     for(int i=0; i<g_mol->numAtoms; i++)  fprintf(checkFile, "%d\n", eachAtomClusterID[i]);
     fclose(checkFile);
 */
     
     for(int i=0; i<g_mol->numAtoms; i++)
         atomListOfBonded[i].clear();
     delete [] atomListOfBonded;
 #endif
 }

◆ buildCrosstermData()

void buildCrosstermData ( )

Definition at line 1165 of file CompressPsf.C.

References crossterm::atom1, crossterm::atom2, crossterm::atom3, crossterm::atom4, crossterm::atom5, crossterm::atom6, crossterm::atom7, crossterm::atom8, CROSSTERM, crossterm::crossterm_type, AtomSigInfo::crosstermSigIndices, eachAtomSigs, g_mol, Molecule::getAllCrossterms(), Molecule::numCrossterms, TupleSignature::setOffsets(), and sigsOfCrossterms.

Referenced by loadMolInfo().

 {
 #ifndef MEM_OPT_VERSION
   Crossterm *crossterms = g_mol->getAllCrossterms();
   //create crossterm's tupleSignature
   for(int i=0; i<g_mol->numCrossterms; i++)
   {
     Crossterm *tuple = crossterms+i;
     TupleSignature oneSig(7, CROSSTERM, tuple->crossterm_type);
     int offset[7];
     offset[0] = tuple->atom2 - tuple->atom1;
     offset[1] = tuple->atom3 - tuple->atom1;
     offset[2] = tuple->atom4 - tuple->atom1;
     offset[3] = tuple->atom5 - tuple->atom1;
     offset[4] = tuple->atom6 - tuple->atom1;
     offset[5] = tuple->atom7 - tuple->atom1;
     offset[6] = tuple->atom8 - tuple->atom1;
     oneSig.setOffsets(offset);
    
     int poolIndex = sigsOfCrossterms.lookupCstPool(oneSig);
     if(poolIndex == -1)
     {
       sigsOfCrossterms.push_back(oneSig);
       poolIndex = (SigIndex)sigsOfCrossterms.size()-1;
     }
     eachAtomSigs[tuple->atom1].crosstermSigIndices.push_back(poolIndex);
   }
   
   delete[] crossterms;
 #endif
 }

◆ buildDihedralData()

void buildDihedralData ( )

Definition at line 1107 of file CompressPsf.C.

References dihedral::atom1, dihedral::atom2, dihedral::atom3, dihedral::atom4, DIHEDRAL, dihedral::dihedral_type, AtomSigInfo::dihedralSigIndices, eachAtomSigs, g_mol, Molecule::getAllDihedrals(), Molecule::numDihedrals, TupleSignature::setOffsets(), and sigsOfDihedrals.

Referenced by loadMolInfo().

 {
 #ifndef MEM_OPT_VERSION
     Dihedral *dihedrals = g_mol->getAllDihedrals();    
 
     //create dihedrals' tupleSignature
     for(int i=0; i<g_mol->numDihedrals; i++)
     {
         Dihedral *tuple = dihedrals+i;
         TupleSignature oneSig(3,DIHEDRAL,tuple->dihedral_type);
         int offset[3];
         offset[0] = tuple->atom2 - tuple->atom1;
         offset[1] = tuple->atom3 - tuple->atom1;
         offset[2] = tuple->atom4 - tuple->atom1;
         oneSig.setOffsets(offset);
 
         int poolIndex = sigsOfDihedrals.lookupCstPool(oneSig);
         if(poolIndex == -1)
         {
             sigsOfDihedrals.push_back(oneSig);
             poolIndex = (SigIndex)sigsOfDihedrals.size()-1;
         }
         eachAtomSigs[tuple->atom1].dihedralSigIndices.push_back(poolIndex);
     }
 
     delete[] dihedrals;
 #endif
 }

◆ buildExclusionData()

void buildExclusionData ( )

◆ buildExclusions()

void buildExclusions ( )

Definition at line 1197 of file CompressPsf.C.

References SimParameters::amberOn, Exclusion::atom1, Exclusion::atom2, atomSigPool, UniqueSetIter< T >::begin(), AtomSigInfo::bondSigIndices, build12Excls(), build13Excls(), build14Excls(), UniqueSet< Elem >::clear(), eachAtomExclSigs, UniqueSetIter< T >::end(), endi(), SimParameters::exclude, ExclSigInfo::fullExclOffset, g_mol, g_simParam, iINFO(), iout, TupleSignature::isReal, ExclSigInfo::modExclOffset, Exclusion::modified, NONE, Molecule::numAtoms, TupleSignature::offset, ONEFOUR, ONETHREE, ONETWO, SimParameters::readExclusions, SCALED14, sigsOfBonds, sigsOfExclusions, UniqueSet< Elem >::size(), and ExclSigInfo::sortExclOffset().

Referenced by compress_molecule_info().

 {
     //1. Build exclusions: mainly accomplish the function of
     //Molecule::build_exclusions (based on the bonds)
     UniqueSet<Exclusion> allExclusions;
 
     int exclude_flag; //Exclusion policy
     exclude_flag = g_simParam->exclude;
     //int stripHGroupExclFlag = (simParams->splitPatch == SPLIT_PATCH_HYDROGEN);
 
     //Commented now since no explicit exclusions are read
     //  Go through the explicit exclusions and add them to the arrays
     //for(i=0; i<numExclusions; i++){
     //  exclusionSet.add(exclusions[i]);
     //}
 
     // If this is AMBER force field, and readExclusions is TRUE,
     // then all the exclusions were read from parm file, and we
     // shouldn't generate any of them.
     // Comment on stripHGroupExcl:
     // 1. Inside this function, hydrogenGroup is initialized in
     // build_atom_status, therefore, not available when reading psf files
     // 2. this function's main purpose is to reduce memory usage. Since exclusion
     // signatures are used, this function could be overlooked  --Chao Mei
 
     vector<int> *eachAtomNeighbors = new vector<int>[g_mol->numAtoms];   
     for(int atom1=0; atom1<g_mol->numAtoms; atom1++)
     {
         AtomSigInfo *aSig = &atomSigPool[atomData[atom1].atomSigIdx];
         for(int j=0; j<aSig->bondSigIndices.size(); j++)
         {
             TupleSignature *tSig = &sigsOfBonds[aSig->bondSigIndices[j]];
             if(!tSig->isReal) continue;
             int atom2 = atom1+tSig->offset[0];
             eachAtomNeighbors[atom1].push_back(atom2);
             eachAtomNeighbors[atom2].push_back(atom1);
         }
     }
 
     if (!g_simParam->amberOn || !g_simParam->readExclusions)
     { //  Now calculate the bonded exlcusions based on the exclusion policy
         switch (exclude_flag)
         {
         case NONE:
             break;
         case ONETWO:
             build12Excls(allExclusions, eachAtomNeighbors);
             break;
         case ONETHREE:
             build12Excls(allExclusions, eachAtomNeighbors);
             build13Excls(allExclusions, eachAtomNeighbors);
             //if ( stripHGroupExclFlag ) stripHGroupExcl();
             break;
         case ONEFOUR:
             build12Excls(allExclusions, eachAtomNeighbors);
             build13Excls(allExclusions, eachAtomNeighbors);
             build14Excls(allExclusions, eachAtomNeighbors, 0);
             //if ( stripHGroupExclFlag ) stripHGroupExcl();
             break;
         case SCALED14:
             build12Excls(allExclusions, eachAtomNeighbors);
             build13Excls(allExclusions, eachAtomNeighbors);
             build14Excls(allExclusions, eachAtomNeighbors, 1);
             //if ( stripHGroupExclFlag ) stripHGroupExcl();
             break;
         }
     }
     //else if (stripHGroupExclFlag && exclude_flag!=NONE && exclude_flag!=ONETWO)
     //  stripHGroupExcl();
 
     //Commented since atomFepFlags information is not available when reading psf file
     //stripFepExcl();
 
     for(int i=0; i<g_mol->numAtoms; i++)
         eachAtomNeighbors[i].clear();
     delete [] eachAtomNeighbors;
 
     //2. Build each atom's list of exclusions
     iout << iINFO << "ADDED " << allExclusions.size() << " IMPLICIT EXCLUSIONS\n" << endi;
     UniqueSetIter<Exclusion> exclIter(allExclusions);
     eachAtomExclSigs = new ExclSigInfo[g_mol->numAtoms];
     for(exclIter=exclIter.begin(); exclIter!=exclIter.end(); exclIter++)
     {
         int atom1 = exclIter->atom1;
         int atom2 = exclIter->atom2;
         int offset21 = atom2-atom1;
         if(exclIter->modified)
         {
             eachAtomExclSigs[atom1].modExclOffset.push_back(offset21);
             eachAtomExclSigs[atom2].modExclOffset.push_back(-offset21);
         }
         else
         {
             eachAtomExclSigs[atom1].fullExclOffset.push_back(offset21);
             eachAtomExclSigs[atom2].fullExclOffset.push_back(-offset21);
         }
     }
     allExclusions.clear();
 
     //3. Build up exclusion signatures and determine each atom's
     //exclusion signature index
     for(int i=0; i<g_mol->numAtoms; i++)
     {
         eachAtomExclSigs[i].sortExclOffset();
         int poolIndex = sigsOfExclusions.lookupCstPool(eachAtomExclSigs[i]);
         if(poolIndex==-1)
         {
             poolIndex = sigsOfExclusions.size();
             sigsOfExclusions.push_back(eachAtomExclSigs[i]);
         }
         atomData[i].exclSigIdx = poolIndex;
     }
     delete [] eachAtomExclSigs;
     eachAtomExclSigs = NULL;
     printf("Exclusion signatures: %d\n", (int)sigsOfExclusions.size());
 }

◆ buildImproperData()

void buildImproperData ( )

Definition at line 1136 of file CompressPsf.C.

References improper::atom1, improper::atom2, improper::atom3, improper::atom4, eachAtomSigs, g_mol, Molecule::getAllImpropers(), IMPROPER, improper::improper_type, AtomSigInfo::improperSigIndices, Molecule::numImpropers, TupleSignature::setOffsets(), and sigsOfImpropers.

Referenced by loadMolInfo().

 { 
 #ifndef MEM_OPT_VERSION
     Improper *impropers=g_mol->getAllImpropers();
 
     //create improper's tupleSignature
     for(int i=0; i<g_mol->numImpropers; i++)
     {
         Improper *tuple = impropers+i;
         TupleSignature oneSig(3,IMPROPER,tuple->improper_type);
         int offset[3];
         offset[0] = tuple->atom2 - tuple->atom1;
         offset[1] = tuple->atom3 - tuple->atom1;
         offset[2] = tuple->atom4 - tuple->atom1;
         oneSig.setOffsets(offset);
 
         int poolIndex = sigsOfImpropers.lookupCstPool(oneSig);
         if(poolIndex == -1)
         {
             sigsOfImpropers.push_back(oneSig);
             poolIndex = (SigIndex)sigsOfImpropers.size()-1;
         }
         eachAtomSigs[tuple->atom1].improperSigIndices.push_back(poolIndex);
     }
 
     delete[] impropers;
 #endif
 }

◆ clearGlobalVectors()

void clearGlobalVectors ( )

Definition at line 419 of file CompressPsf.C.

References atomNamePool, atomTypePool, chargePool, eachClusterSize, massPool, resNamePool, segNamePool, sigsOfAngles, sigsOfBonds, sigsOfDihedrals, sigsOfExclusions, and sigsOfImpropers.

 {
     segNamePool.clear();
     resNamePool.clear();
     atomNamePool.clear();
     atomTypePool.clear();
     chargePool.clear();
     massPool.clear();
     sigsOfBonds.clear();
     sigsOfAngles.clear();
     sigsOfDihedrals.clear();
     sigsOfImpropers.clear();
     sigsOfExclusions.clear();
 
     eachClusterSize.clear();
 }

◆ compress_molecule_info()

void compress_molecule_info	(	Molecule *	mol,
		char *	psfFileName,
		Parameters *	param,
		SimParameters *	simParam,
		ConfigList *	cfgList
	)

Definition at line 436 of file CompressPsf.C.

References buildExclusions(), g_cfgList, g_mol, g_param, g_simParam, integrateAllAtomSigs(), loadMolInfo(), and outputCompressedFile().

Referenced by NamdState::loadStructure().

 {
     g_mol = mol;
     g_param = param;
     g_simParam = simParam; //used for building exclusions
     g_cfgList = cfgList; //used for integrating extra bonds
 
     //read psf files
     //readPsfFile(psfFileName);
     loadMolInfo();
 
     integrateAllAtomSigs();
 
     buildExclusions();
 
     //buildParamData();
 
     char *outFileName = new char[strlen(psfFileName)+20];
     sprintf(outFileName, "%s.inter", psfFileName);
     //the text file for signatures and other non-per-atom info
     FILE *txtOfp = fopen(outFileName, "w");
     sprintf(outFileName, "%s.inter.bin", psfFileName);
     //the binary file for per-atom info
     FILE *binOfp = fopen(outFileName, "wb");
     delete [] outFileName;
 
     //output compressed psf file
     outputCompressedFile(txtOfp, binOfp);
 
     fclose(txtOfp);
     fclose(binOfp);
 }

◆ flipNum()

void flipNum	(	char *	elem,
		int	elemSize,
		int	numElems
	)

Definition at line 406 of file CompressPsf.C.

Referenced by OutputAtomRecord::flip().

                                                     {
     int mid = elemSize/2;
     char *ptr = elem;
     for(int i=0; i<numElems; i++) {
         for(int j=0; j<mid; j++) {
             char tmp = ptr[j];
             ptr[j] = ptr[elemSize-1-j];
             ptr[elemSize-1-j] = tmp;
         }
         ptr += elemSize;
     }
 }

◆ getExtraBonds()

void getExtraBonds ( StringList * file )

◆ integrateAllAtomSigs()

void integrateAllAtomSigs ( )

Definition at line 505 of file CompressPsf.C.

References atomSigPool, eachAtomSigs, g_mol, Molecule::numAtoms, sigsOfAngles, sigsOfBonds, sigsOfCrossterms, sigsOfDihedrals, sigsOfImpropers, and AtomSigInfo::sortTupleSigIndices().

Referenced by compress_molecule_info().

 {
     printf("Bond sigs:  %d\n", (int)sigsOfBonds.size());
     printf("Angle sigs:  %d\n", (int)sigsOfAngles.size());
     printf("Dihedral sigs:  %d\n", (int)sigsOfDihedrals.size());
     printf("Improper sigs:  %d\n", (int)sigsOfImpropers.size());
     printf("Crossterm sigs:  %d\n", (int)sigsOfCrossterms.size());
 
 
     for(int i=0; i<g_mol->numAtoms; i++)
     {
         eachAtomSigs[i].sortTupleSigIndices();
         int poolIndex = atomSigPool.lookupCstPool(eachAtomSigs[i]);
         if(poolIndex==-1)
         {
             atomSigPool.push_back(eachAtomSigs[i]);
             poolIndex = atomSigPool.size()-1;
         }
         atomData[i].atomSigIdx = poolIndex;
     }
 
     printf("Atom's sigs: %d\n", (int)atomSigPool.size());
 
     delete[] eachAtomSigs;
 }

◆ loadMolInfo()

void loadMolInfo ( )

Before entering this function, all information about Molecule has been obtained. The Molecule::build_atom_status should also be called.

Definition at line 473 of file CompressPsf.C.

References buildAngleData(), buildAtomData(), buildBondData(), buildCrosstermData(), buildDihedralData(), buildImproperData(), eachAtomSigs, g_mol, g_param, and Molecule::numAtoms.

Referenced by compress_molecule_info().

 {
     char err_msg[512];  //  Error message for NAMD_die
     char buffer[512];  //  Buffer for file reading
     int i;      //  Loop counter
     int NumTitle;    //  Number of Title lines in .psf file
     int ret_code;    //  ret_code from NAMD_read_line calls
     FILE *psf_file;
     Parameters *params = g_param;
 
     buildAtomData();
 
     //read extra bonds/angles/dihedrals/impropers information first
     //and then integrate them into the following reading procedures
     /*if(g_simParam->extraBondsOn)
     {
         getExtraBonds(g_cfgList->find("extraBondsFile"));
     }*/
 
     //initialize eachAtomSigs
     eachAtomSigs = new AtomSigInfo[g_mol->numAtoms];    
 
     buildBondData();
     buildAngleData();
     buildDihedralData();
     buildImproperData();
     buildCrosstermData();
 
     //  analyze the data and find the status of each atom
     //build_atom_status();
 }

◆ operator!=()

int operator!=	(	const FourBodyConsts &	f1,
		const FourBodyConsts &	f2
	)

Definition at line 353 of file CompressPsf.C.

References four_body_consts::delta, four_body_consts::k, and four_body_consts::n.

 {
     return (f1.delta!=f2.delta) || (f1.k!=f2.k) || (f1.n!=f2.n);
 }

◆ operator==() [1/6]

int operator==	(	const AtomSigInfo &	s1,
		const AtomSigInfo &	s2
	)

Definition at line 146 of file CompressPsf.C.

References AtomSigInfo::angleSigIndices, AtomSigInfo::bondSigIndices, AtomSigInfo::crosstermSigIndices, AtomSigInfo::dihedralSigIndices, and AtomSigInfo::improperSigIndices.

 {
     if(s1.bondSigIndices.size() != s2.bondSigIndices.size())
         return 0;
     if(s1.angleSigIndices.size() != s2.angleSigIndices.size())
         return 0;
     if(s1.dihedralSigIndices.size() != s2.dihedralSigIndices.size())
         return 0;
     if(s1.improperSigIndices.size() != s2.improperSigIndices.size())
         return 0;
     if(s1.crosstermSigIndices.size() != s2.crosstermSigIndices.size())
         return 0;
 
     int equalCnt;
     equalCnt=0;
     int bondSigCnt = s1.bondSigIndices.size();
     for(int i=0; i<bondSigCnt; i++)
         equalCnt += (s1.bondSigIndices[i]==s2.bondSigIndices[i]);
     if(equalCnt!=bondSigCnt)
         return 0;
 
     equalCnt=0;
     int angleSigCnt = s1.angleSigIndices.size();
     for(int i=0; i<angleSigCnt; i++)
         equalCnt += (s1.angleSigIndices[i]==s2.angleSigIndices[i]);
     if(equalCnt!=angleSigCnt)
         return 0;
 
     equalCnt=0;
     int dihedralSigCnt = s1.dihedralSigIndices.size();
     for(int i=0; i<dihedralSigCnt; i++)
         equalCnt += (s1.dihedralSigIndices[i]==s2.dihedralSigIndices[i]);
     if(equalCnt!=dihedralSigCnt)
         return 0;
 
     equalCnt=0;
     int improperSigCnt = s1.improperSigIndices.size();
     for(int i=0; i<improperSigCnt; i++)
         equalCnt += (s1.improperSigIndices[i]==s2.improperSigIndices[i]);
     if(equalCnt!=improperSigCnt)
         return 0;
 
     equalCnt=0;
     int crosstermSigCnt = s1.crosstermSigIndices.size();
     for(int i=0; i<crosstermSigCnt; i++)
         equalCnt += (s1.crosstermSigIndices[i]==s2.crosstermSigIndices[i]);
     if(equalCnt!=crosstermSigCnt)
         return 0;
 
     return 1;
 }

◆ operator==() [2/6]

int operator==	(	const ExclSigInfo &	s1,
		const ExclSigInfo &	s2
	)

Definition at line 253 of file CompressPsf.C.

References ExclSigInfo::fullExclOffset, and ExclSigInfo::modExclOffset.

 {
     if(s1.fullExclOffset.size()!=s2.fullExclOffset.size())
         return 0;
     if(s1.modExclOffset.size()!=s2.modExclOffset.size())
         return 0;
 
     for(int i=0; i<s1.fullExclOffset.size(); i++)
     {
         if(s1.fullExclOffset[i] != s2.fullExclOffset[i])
             return 0;
     }
 
     for(int i=0; i<s1.modExclOffset.size(); i++)
     {
         if(s1.modExclOffset[i] != s2.modExclOffset[i])
             return 0;
     }
     return 1;
 }

◆ operator==() [3/6]

int operator==	(	const BondValue &	b1,
		const BondValue &	b2
	)

Definition at line 343 of file CompressPsf.C.

References BondValue::k, and BondValue::x0.

 {
     return (b1.k==b2.k) && (b1.x0==b2.x0);
 }

◆ operator==() [4/6]

int operator==	(	const AngleValue &	a1,
		const AngleValue &	a2
	)

Definition at line 348 of file CompressPsf.C.

References AngleValue::k, AngleValue::k_ub, AngleValue::r_ub, and AngleValue::theta0.

 {
     return (a1.k==a2.k) && (a1.k_ub==a2.k_ub) && (a1.r_ub==a2.r_ub) && (a1.theta0==a2.theta0);
 }

◆ operator==() [5/6]

int operator==	(	const DihedralValue &	d1,
		const DihedralValue &	d2
	)

Definition at line 358 of file CompressPsf.C.

References MAX_MULTIPLICITY, DihedralValue::multiplicity, and DihedralValue::values.

 {
     if(d1.multiplicity != d2.multiplicity)
         return 0;
     for(int i=0; i<MAX_MULTIPLICITY; i++)
     {
         if(d1.values[i] != d2.values[i])
             return 0;
     }
     return 1;
 }

◆ operator==() [6/6]

int operator==	(	const ImproperValue &	d1,
		const ImproperValue &	d2
	)

Definition at line 370 of file CompressPsf.C.

References MAX_MULTIPLICITY, ImproperValue::multiplicity, and ImproperValue::values.

 {
     if(d1.multiplicity != d2.multiplicity)
         return 0;
     for(int i=0; i<MAX_MULTIPLICITY; i++)
     {
         if(d1.values[i] != d2.values[i])
             return 0;
     }
     return 1;
 }

◆ outputCompressedFile()

void outputCompressedFile	(	FILE *	txtOfp,
		FILE *	binOfp
	)

Output the compressed psf files. The binary per-atom file contains two part. The first part is used for the parallel input containing info such as atom signature ids; the second part is used for the parallel output containing infor such as cluster ids and whether an atom is water or not.

-Chao Mei

Definition at line 541 of file CompressPsf.C.

Referenced by compress_molecule_info().

 {
 #ifndef MEM_OPT_VERSION
     fprintf(txtOfp, "FORMAT VERSION: %f\n", COMPRESSED_PSF_VER);
 
     fprintf(txtOfp, "%d !NSEGMENTNAMES\n", segNamePool.size());
     for(int i=0; i<segNamePool.size(); i++)
     {
         fprintf(txtOfp, "%s\n", segNamePool[i].c_str());
     }
 
     fprintf(txtOfp, "%d !NRESIDUENAMES\n", resNamePool.size());
     for(int i=0; i<resNamePool.size(); i++)
     {
         fprintf(txtOfp, "%s\n", resNamePool[i].c_str());
     }
 
     fprintf(txtOfp, "%d !NATOMNAMES\n", atomNamePool.size());
     for(int i=0; i<atomNamePool.size(); i++)
     {
         fprintf(txtOfp, "%s\n", atomNamePool[i].c_str());
     }
 
     fprintf(txtOfp, "%d !NATOMTYPES\n", atomTypePool.size());
     for(int i=0; i<atomTypePool.size(); i++)
     {
         fprintf(txtOfp, "%s\n", atomTypePool[i].c_str());
     }
 
     fprintf(txtOfp, "%d !NCHARGES\n", chargePool.size());
     for(int i=0; i<chargePool.size(); i++)
     {
         const Real charge = chargePool[i];
         fprintf(txtOfp, "%f\n", charge);
     }
 
     fprintf(txtOfp, "%d !NMASSES\n", massPool.size());
     for(int i=0; i<massPool.size(); i++)
     {
         const Real mass = massPool[i];
         fprintf(txtOfp, "%f\n", mass);
     }
 
 
     fprintf(txtOfp, "%d !NATOMSIGS\n", atomSigPool.size());
     for(int i=0; i<atomSigPool.size(); i++)
     {
         AtomSigInfo& oneAtomSig = atomSigPool[i];
         int oneTypeCnt = oneAtomSig.bondSigIndices.size();
         fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NBOND");
         for(int j=0; j<oneTypeCnt; j++)
         {
             SigIndex idx = oneAtomSig.bondSigIndices[j];
             TupleSignature& tSig = sigsOfBonds[idx];
             tSig.output(txtOfp);
         }
 
         oneTypeCnt = oneAtomSig.angleSigIndices.size();
         fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NTHETA");
         for(int j=0; j<oneTypeCnt; j++)
         {
             SigIndex idx = oneAtomSig.angleSigIndices[j];
             TupleSignature& tSig = sigsOfAngles[idx];
             tSig.output(txtOfp);
         }
 
         oneTypeCnt = oneAtomSig.dihedralSigIndices.size();
         fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NPHI");
         for(int j=0; j<oneTypeCnt; j++)
         {
             SigIndex idx = oneAtomSig.dihedralSigIndices[j];
             TupleSignature& tSig = sigsOfDihedrals[idx];
             tSig.output(txtOfp);
         }
 
         oneTypeCnt = oneAtomSig.improperSigIndices.size();
         fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NIMPHI");
         for(int j=0; j<oneTypeCnt; j++)
         {
             SigIndex idx = oneAtomSig.improperSigIndices[j];
             TupleSignature& tSig = sigsOfImpropers[idx];
             tSig.output(txtOfp);
         }
 
         oneTypeCnt = oneAtomSig.crosstermSigIndices.size();
         fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NCRTERM");
         for(int j=0; j<oneTypeCnt; j++)
         {
             SigIndex idx = oneAtomSig.crosstermSigIndices[j];
             TupleSignature& tSig = sigsOfCrossterms[idx];
             tSig.output(txtOfp);
         }
     }
 
     //2. Output exclusion signatures
     int exclSigCnt = sigsOfExclusions.size();
     fprintf(txtOfp, "%d !NEXCLSIGS\n", exclSigCnt);
     for(int i=0; i<exclSigCnt; i++)
     {
         ExclSigInfo *sig = &sigsOfExclusions[i];
         //first line is for full exclusions (1-2, 1-3) in the format of count offset1 offset2 offset3 ...
         fprintf(txtOfp, "%lu", (unsigned long) sig->fullExclOffset.size());
         for(int j=0; j<sig->fullExclOffset.size(); j++)
             fprintf(txtOfp, " %d", sig->fullExclOffset[j]);
         fprintf(txtOfp, "\n");
 
         //second line is for modified exclusions (1-4)
         fprintf(txtOfp, "%lu", (unsigned long) sig->modExclOffset.size());
         for(int j=0; j<sig->modExclOffset.size(); j++)
             fprintf(txtOfp, " %d", sig->modExclOffset[j]);
         fprintf(txtOfp, "\n");
     }
 
     //3. Output the cluster information
     fprintf(txtOfp, "%d !NCLUSTERS\n", g_numClusters);
 
     //4. Output atom info
     fprintf(txtOfp, "%d !NATOM\n", g_mol->numAtoms);
     fprintf(txtOfp, "%d !NHYDROGENGROUP\n", g_mol->numHydrogenGroups);
     fprintf(txtOfp, "%d !MAXHYDROGENGROUPSIZE\n", g_mol->maxHydrogenGroupSize);
     fprintf(txtOfp, "%d !NMIGRATIONGROUP\n", g_mol->numMigrationGroups);
     fprintf(txtOfp, "%d !MAXMIGRATIONGROUPSIZE\n", g_mol->maxMigrationGroupSize);
 
     //5. Output rigid bond type
     fprintf(txtOfp, "%d !RIGIDBONDTYPE\n", g_simParam->rigidBonds);
 #if 0
     const float *atomOccupancy = g_mol->getOccupancyData();
     const float *atomBFactor = g_mol->getBFactorData();
     fprintf(txtOfp, "%d !OCCUPANCYVALID\n", (atomOccupancy==NULL)?0:1);
     fprintf(txtOfp, "%d !TEMPFACTORVALID\n", (atomBFactor==NULL)?0:1);
 
     float *zeroFloats = NULL;
     if(atomOccupancy==NULL || atomBFactor==NULL) {
         zeroFloats = new float[g_mol->numAtoms];
         memset(zeroFloats, 0, sizeof(float)*g_mol->numAtoms);
         if(atomOccupancy==NULL) atomOccupancy = (const float *)zeroFloats;
         if(atomBFactor==NULL) atomBFactor = (const float *)zeroFloats;
     }
 #endif
 
     Atom *atoms = g_mol->getAtoms(); //need to output its partner and hydrogenList
     HydrogenGroupID *hg = g_mol->hydrogenGroup.begin();
 
     //First, output magic number
     int magicNum = COMPRESSED_PSF_MAGICNUM;
     fwrite(&magicNum, sizeof(int), 1, binOfp);
     //Second, version number
     float verNum = (float)COMPRESSED_PSF_VER;
     fwrite(&verNum, sizeof(float), 1, binOfp);
     //Third, the per-atom record size
     int recSize = sizeof(OutputAtomRecord);
     fwrite(&recSize, sizeof(int), 1, binOfp);
     //Fourth, each atom info
     OutputAtomRecord oneRec;
     for(int i=0; i<g_mol->numAtoms; i++)
     {                
         oneRec.sSet.segNameIdx = atomData[i].segNameIdx;
         oneRec.sSet.resNameIdx = atomData[i].resNameIdx;
         oneRec.sSet.atomNameIdx = atomData[i].atomNameIdx;
         oneRec.sSet.atomTypeIdx = atomData[i].atomTypeIdx;
         oneRec.sSet.chargeIdx = atomData[i].chargeIdx;
         oneRec.sSet.massIdx = atomData[i].massIdx;
         oneRec.iSet.atomSigIdx = atomData[i].atomSigIdx;
 
         oneRec.iSet.exclSigIdx = atomData[i].exclSigIdx;        
         oneRec.sSet.vdw_type = atoms[i].vdw_type;
         oneRec.iSet.resID = atomData[i].resID;        
         int hydIdx = atoms[i].hydrogenList;       
         oneRec.iSet.hydrogenList = hydIdx;
         oneRec.iSet.atomsInGroup = hg[hydIdx].atomsInGroup;
         oneRec.iSet.GPID = hg[hydIdx].GPID;
         //oneRec.waterVal = hg[hydIdx].waterVal;
         oneRec.iSet.atomsInMigrationGroup = hg[hydIdx].atomsInMigrationGroup;
         oneRec.iSet.MPID = hg[hydIdx].MPID;
         //oneRec.fSet.occupancy = atomOccupancy[i];
         //oneRec.fSet.bfactor = atomBFactor[i];
         oneRec.fSet.rigidBondLength = g_mol->rigid_bond_length(i);
         fwrite(&oneRec, sizeof(OutputAtomRecord), 1, binOfp);
     }
     //if(zeroFloats) delete[] zeroFloats;
     delete[] atomData;
 
     //Output the info required for parallel output:
     //1. Cluster ids of each atom
     //Since the cluster info is only when doing output under
     //wrapAll or wrapWater conditions, for the purpose of parallel
     //output, it is reasonable to separate the cluster info from
     //the above per-atom info. So whole the binary per-atom file
     //contains two parts, one for parallel input to create patches
     //and computes; the other for parallel output -Chao Mei
     //Fifth: output the cluster id of each atoms
     fwrite(eachAtomClusterID, sizeof(int), g_mol->numAtoms, binOfp);    
     //2. Whether each atom is water or not
     char *isWater = new char[g_mol->numAtoms];
     for(int i=0; i<g_mol->numAtoms; i++){
         isWater[i] = g_mol->is_water(i);
     }
     fwrite(isWater, sizeof(char), g_mol->numAtoms, binOfp);
     delete [] isWater;
     delete[] atoms;
     g_mol->hydrogenGroup.resize(0);
     delete[] eachAtomClusterID;    
 
     //Output the parameter new values if extraBonds are present.
     //The parameters are not needed since now extraBonds' parameters will be
     //read again during running the simulation
 
     //6. Output the "multiplicity" field TUPLE_array of the Parameter object
     fprintf(txtOfp, "!DIHEDRALPARAMARRAY\n");
     for(int i=0; i<g_param->NumDihedralParams; i++)
     {
         fprintf(txtOfp, "%d ", g_param->dihedral_array[i].multiplicity);
     }
     fprintf(txtOfp, "\n");
     fprintf(txtOfp, "!IMPROPERPARAMARRAY\n");
     for(int i=0; i<g_param->NumImproperParams; i++)
     {
         fprintf(txtOfp, "%d ", g_param->improper_array[i].multiplicity);
     }
     fprintf(txtOfp, "\n");
 #endif
 }

Variable Documentation

◆ atomData

BasicAtomInfo* atomData

Definition at line 314 of file CompressPsf.C.

Referenced by outputCompressedFile().

◆ atomNamePool

HashPool<HashString> atomNamePool

Definition at line 309 of file CompressPsf.C.

Referenced by buildAtomData(), clearGlobalVectors(), Molecule::get_atom_from_name(), outputCompressedFile(), Molecule::receive_Molecule(), and Molecule::send_Molecule().

◆ atomSigPool

HashPool<AtomSigInfo> atomSigPool

Definition at line 313 of file CompressPsf.C.

Referenced by buildExclusions(), integrateAllAtomSigs(), outputCompressedFile(), Molecule::receive_Molecule(), Molecule::send_Molecule(), ParallelIOMgr::updateMolInfo(), and Molecule::~Molecule().

◆ atomTypePool

HashPool<HashString> atomTypePool

Definition at line 310 of file CompressPsf.C.

Referenced by buildAtomData(), clearGlobalVectors(), Molecule::get_atomtype(), and outputCompressedFile().

◆ chargePool

HashPool<HashReal> chargePool

Definition at line 311 of file CompressPsf.C.

Referenced by buildAtomData(), clearGlobalVectors(), and outputCompressedFile().

◆ eachAtomClusterID

int* eachAtomClusterID = NULL

Definition at line 317 of file CompressPsf.C.

Referenced by buildBondData(), and outputCompressedFile().

◆ eachAtomExclSigs

ExclSigInfo* eachAtomExclSigs

Definition at line 330 of file CompressPsf.C.

Referenced by buildExclusions().

◆ eachAtomSigs

AtomSigInfo* eachAtomSigs

Definition at line 327 of file CompressPsf.C.

Referenced by buildAngleData(), buildBondData(), buildCrosstermData(), buildDihedralData(), buildImproperData(), integrateAllAtomSigs(), and loadMolInfo().

◆ eachClusterID

vector<int> eachClusterID

Definition at line 319 of file CompressPsf.C.

Referenced by buildBondData().

◆ eachClusterSize

vector<int> eachClusterSize

Definition at line 318 of file CompressPsf.C.

Referenced by buildBondData(), and clearGlobalVectors().

◆ extraAngleParams

vector<AngleValue> extraAngleParams

Definition at line 339 of file CompressPsf.C.

◆ extraAngles

vector<Angle> extraAngles

Definition at line 334 of file CompressPsf.C.

◆ extraBondParams

vector<BondValue> extraBondParams

Definition at line 338 of file CompressPsf.C.

◆ extraBonds

vector<Bond> extraBonds

Definition at line 333 of file CompressPsf.C.

◆ extraDihedralParams

vector<DihedralValue> extraDihedralParams

Definition at line 340 of file CompressPsf.C.

◆ extraDihedrals

vector<Dihedral> extraDihedrals

Definition at line 335 of file CompressPsf.C.

◆ extraImproperParams

vector<ImproperValue> extraImproperParams

Definition at line 341 of file CompressPsf.C.

◆ extraImpropers

vector<Improper> extraImpropers

Definition at line 336 of file CompressPsf.C.

◆ g_cfgList

ConfigList* g_cfgList = NULL

Definition at line 35 of file CompressPsf.C.

Referenced by compress_molecule_info().

◆ g_mol

Molecule* g_mol = NULL

Very tricky part:

Generating the tuple type affects the Parameter object. Particularly, the "multiplicity" field in the TUPLE_array will be changed in the function assign_TUPLE_index. TUPLE=dihedral, improper.

And when we read the compressed psf files, assign_TUPLE_index functions will not be called so that the value of "multiplicity" will not be updated. This results in the incorrect energy from bonded interaction.

Therefore, we need to store its value in the compressed psf file. When it comes to read them, we need to update the Parameter object with these values.

Definition at line 32 of file CompressPsf.C.

Referenced by build12Excls(), build13Excls(), build14Excls(), buildAngleData(), buildAtomData(), buildBondData(), buildCrosstermData(), buildDihedralData(), buildExclusions(), buildImproperData(), compress_molecule_info(), integrateAllAtomSigs(), loadMolInfo(), and outputCompressedFile().