ViewVC Help

View File | Revision Log | Show Annotations | View Changeset | Root Listing

root/OpenMD/trunk/src/brains/SimCreator.cpp

Comparing trunk/src/brains/SimCreator.cpp (file contents):
Revision 1025 by gezelter, Wed Aug 30 20:33:44 2006 UTC vs.
Revision 1993 by gezelter, Tue Apr 29 17:32:31 2014 UTC

#	Line 6 | Line 6
6		* redistribute this software in source and binary code form, provided
7		* that the following conditions are met:
8		*
9	<	* 1. Acknowledgement of the program authors must be made in any
10	<	*    publication of scientific results based in part on use of the
11	<	*    program.  An acceptable form of acknowledgement is citation of
12	<	*    the article in which the program was described (Matthew
13	<	*    A. Meineke, Charles F. Vardeman II, Teng Lin, Christopher
14	<	*    J. Fennell and J. Daniel Gezelter, "OOPSE: An Object-Oriented
15	<	*    Parallel Simulation Engine for Molecular Dynamics,"
16	<	*    J. Comput. Chem. 26, pp. 252-271 (2005))
17	<	*
18	<	* 2. Redistributions of source code must retain the above copyright
9	>	* 1. Redistributions of source code must retain the above copyright
10		*    notice, this list of conditions and the following disclaimer.
11		*
12	<	* 3. Redistributions in binary form must reproduce the above copyright
12	>	* 2. Redistributions in binary form must reproduce the above copyright
13		*    notice, this list of conditions and the following disclaimer in the
14		*    documentation and/or other materials provided with the
15		*    distribution.
#	Line 37 | Line 28
28		* arising out of the use of or inability to use software, even if the
29		* University of Notre Dame has been advised of the possibility of
30		* such damages.
31	+	*
32	+	* SUPPORT OPEN SCIENCE!  If you use OpenMD or its source code in your
33	+	* research, please cite the appropriate papers when you publish your
34	+	* work.  Good starting points are:
35	+	*
36	+	* [1]  Meineke, et al., J. Comp. Chem. 26, 252-271 (2005).
37	+	* [2]  Fennell & Gezelter, J. Chem. Phys. 124, 234104 (2006).
38	+	* [3]  Sun, Lin & Gezelter, J. Chem. Phys. 128, 234107 (2008).
39	+	* [4]  Kuang & Gezelter,  J. Chem. Phys. 133, 164101 (2010).
40	+	* [5]  Vardeman, Stocker & Gezelter, J. Chem. Theory Comput. 7, 834 (2011).
41		*/
42
43		/**
44		* @file SimCreator.cpp
45		* @author tlin
46		* @date 11/03/2004
46	–	* @time 13:51am
47		* @version 1.0
48		*/
49	+
50	+	#ifdef IS_MPI
51	+	#include "mpi.h"
52	+	#include "math/ParallelRandNumGen.hpp"
53	+	#endif
54	+
55		#include <exception>
56		#include <iostream>
57		#include <sstream>
#	Line 55 | Line 61
61		#include "brains/SimCreator.hpp"
62		#include "brains/SimSnapshotManager.hpp"
63		#include "io/DumpReader.hpp"
64	<	#include "UseTheForce/ForceFieldFactory.hpp"
64	>	#include "brains/ForceField.hpp"
65		#include "utils/simError.h"
66		#include "utils/StringUtils.hpp"
67	+	#include "utils/Revision.hpp"
68		#include "math/SeqRandNumGen.hpp"
69		#include "mdParser/MDLexer.hpp"
70		#include "mdParser/MDParser.hpp"
#	Line 75 | Line 82
82		#include "antlr/NoViableAltForCharException.hpp"
83		#include "antlr/NoViableAltException.hpp"
84
85	<	#ifdef IS_MPI
86	<	#include "math/ParallelRandNumGen.hpp"
87	<	#endif
85	>	#include "types/DirectionalAdapter.hpp"
86	>	#include "types/MultipoleAdapter.hpp"
87	>	#include "types/EAMAdapter.hpp"
88	>	#include "types/SuttonChenAdapter.hpp"
89	>	#include "types/PolarizableAdapter.hpp"
90	>	#include "types/FixedChargeAdapter.hpp"
91	>	#include "types/FluctuatingChargeAdapter.hpp"
92
93	<	namespace oopse {
93	>
94	>	namespace OpenMD {
95
96	<	Globals* SimCreator::parseFile(std::istream& rawMetaDataStream, const std::string& filename, int startOfMetaDataBlock ){
96	>	Globals* SimCreator::parseFile(std::istream& rawMetaDataStream, const std::string& filename, int mdFileVersion, int startOfMetaDataBlock ){
97		Globals* simParams = NULL;
98		try {
99
#	Line 90 | Line 102 | namespace oopse {
102		#ifdef IS_MPI
103		int streamSize;
104		const int masterNode = 0;
105	<	int commStatus;
105	>
106		if (worldRank == masterNode) {
107	<	#endif
108	<
107	>	MPI_Bcast(&mdFileVersion, 1, MPI_INT, masterNode, MPI_COMM_WORLD);
108	>	#endif
109		SimplePreprocessor preprocessor;
110	<	preprocessor.preprocess(rawMetaDataStream, filename, startOfMetaDataBlock, ppStream);
110	>	preprocessor.preprocess(rawMetaDataStream, filename,
111	>	startOfMetaDataBlock, ppStream);
112
113		#ifdef IS_MPI
114	<	//brocasting the stream size
114	>	//broadcasting the stream size
115		streamSize = ppStream.str().size() +1;
116	<	commStatus = MPI_Bcast(&streamSize, 1, MPI_LONG, masterNode, MPI_COMM_WORLD);
117	<
118	<	commStatus = MPI_Bcast(static_cast<void*>(const_cast<char*>(ppStream.str().c_str())), streamSize, MPI_CHAR, masterNode, MPI_COMM_WORLD);
106	<
107	<
116	>	MPI_Bcast(&streamSize, 1, MPI_INT, masterNode, MPI_COMM_WORLD);
117	>	MPI_Bcast(static_cast<void*>(const_cast<char*>(ppStream.str().c_str())),
118	>	streamSize, MPI_CHAR, masterNode, MPI_COMM_WORLD);
119		} else {
120	+
121	+	MPI_Bcast(&mdFileVersion, 1, MPI_INT, masterNode, MPI_COMM_WORLD);
122	+
123		//get stream size
124	<	commStatus = MPI_Bcast(&streamSize, 1, MPI_LONG, masterNode, MPI_COMM_WORLD);
111	<
124	>	MPI_Bcast(&streamSize, 1, MPI_INT, masterNode, MPI_COMM_WORLD);
125		char* buf = new char[streamSize];
126		assert(buf);
127
128		//receive file content
129	<	commStatus = MPI_Bcast(buf, streamSize, MPI_CHAR, masterNode, MPI_COMM_WORLD);
117	<
118	<	ppStream.str(buf);
119	<	delete buf;
129	>	MPI_Bcast(buf, streamSize, MPI_CHAR, masterNode, MPI_COMM_WORLD);
130
131	+	ppStream.str(buf);
132	+	delete [] buf;
133		}
134		#endif
135		// Create a scanner that reads from the input stream
#	Line 139 | Line 151 | namespace oopse {
151		parser.initializeASTFactory(factory);
152		parser.setASTFactory(&factory);
153		parser.mdfile();
142	–
154		// Create a tree parser that reads information into Globals
155		MDTreeParser treeParser;
156		treeParser.initializeASTFactory(factory);
157		treeParser.setASTFactory(&factory);
158		simParams = treeParser.walkTree(parser.getAST());
148	–
159		}
160
161
#	Line 215 | Line 225 | namespace oopse {
225		painCave.isFatal = 1;
226		simError();
227		}
228	<	catch (OOPSEException& e) {
228	>	catch (OpenMDException& e) {
229		sprintf(painCave.errMsg,
230		"%s\n",
231		e.getMessage().c_str());
#	Line 230 | Line 240 | namespace oopse {
240		simError();
241		}
242
243	+	simParams->setMDfileVersion(mdFileVersion);
244		return simParams;
245		}
246
247		SimInfo*  SimCreator::createSim(const std::string & mdFileName,
248		bool loadInitCoords) {
249	<
249	>
250		const int bufferSize = 65535;
251		char buffer[bufferSize];
252		int lineNo = 0;
253		std::string mdRawData;
254		int metaDataBlockStart = -1;
255		int metaDataBlockEnd = -1;
256	<	int i;
257	<	int mdOffset;
256	>	int i, j;
257	>	streamoff mdOffset;
258	>	int mdFileVersion;
259
260	+	// Create a string for embedding the version information in the MetaData
261	+	std::string version;
262	+	version.assign("## Last run using OpenMD Version: ");
263	+	version.append(OPENMD_VERSION_MAJOR);
264	+	version.append(".");
265	+	version.append(OPENMD_VERSION_MINOR);
266	+
267	+	std::string svnrev(g_REVISION, strnlen(g_REVISION, 20));
268	+	//convert a macro from compiler to a string in c++
269	+	// STR_DEFINE(svnrev, SVN_REV );
270	+	version.append(" Revision: ");
271	+	// If there's no SVN revision, just call this the RELEASE revision.
272	+	if (!svnrev.empty()) {
273	+	version.append(svnrev);
274	+	} else {
275	+	version.append("RELEASE");
276	+	}
277	+
278		#ifdef IS_MPI
279		const int masterNode = 0;
280		if (worldRank == masterNode) {
281		#endif
282
283	<	std::ifstream mdFile_(mdFileName.c_str());
283	>	std::ifstream mdFile_;
284	>	mdFile_.open(mdFileName.c_str(), ifstream::in | ifstream::binary);
285
286		if (mdFile_.fail()) {
287		sprintf(painCave.errMsg,
#	Line 263 | Line 294 | namespace oopse {
294		mdFile_.getline(buffer, bufferSize);
295		++lineNo;
296		std::string line = trimLeftCopy(buffer);
297	<	i = CaseInsensitiveFind(line, "<OOPSE");
298	<	if (i == string::npos) {
297	>	i = CaseInsensitiveFind(line, "<OpenMD");
298	>	if (static_cast<size_t>(i) == string::npos) {
299	>	// try the older file strings to see if that works:
300	>	i = CaseInsensitiveFind(line, "<OOPSE");
301	>	}
302	>
303	>	if (static_cast<size_t>(i) == string::npos) {
304	>	// still no luck!
305		sprintf(painCave.errMsg,
306	<	"SimCreator: File: %s is not an OOPSE file!\n",
306	>	"SimCreator: File: %s is not a valid OpenMD file!\n",
307		mdFileName.c_str());
308		painCave.isFatal = 1;
309		simError();
310		}
311	+
312	+	// found the correct opening string, now try to get the file
313	+	// format version number.
314
315	+	StringTokenizer tokenizer(line, "=<> \t\n\r");
316	+	std::string fileType = tokenizer.nextToken();
317	+	toUpper(fileType);
318	+
319	+	mdFileVersion = 0;
320	+
321	+	if (fileType == "OPENMD") {
322	+	while (tokenizer.hasMoreTokens()) {
323	+	std::string token(tokenizer.nextToken());
324	+	toUpper(token);
325	+	if (token == "VERSION") {
326	+	mdFileVersion = tokenizer.nextTokenAsInt();
327	+	break;
328	+	}
329	+	}
330	+	}
331	+
332		//scan through the input stream and find MetaData tag
333		while(mdFile_.getline(buffer, bufferSize)) {
334		++lineNo;
#	Line 312 | Line 369 | namespace oopse {
369
370		mdRawData.clear();
371
372	+	bool foundVersion = false;
373	+
374		for (int i = 0; i < metaDataBlockEnd - metaDataBlockStart - 1; ++i) {
375		mdFile_.getline(buffer, bufferSize);
376	<	mdRawData += buffer;
376	>	std::string line = trimLeftCopy(buffer);
377	>	j = CaseInsensitiveFind(line, "## Last run using OpenMD Version");
378	>	if (static_cast<size_t>(j) != string::npos) {
379	>	foundVersion = true;
380	>	mdRawData += version;
381	>	} else {
382	>	mdRawData += buffer;
383	>	}
384		mdRawData += "\n";
385		}
386	<
386	>
387	>	if (!foundVersion) mdRawData += version + "\n";
388	>
389		mdFile_.close();
390
391		#ifdef IS_MPI
#	Line 327 | Line 395 | namespace oopse {
395		std::stringstream rawMetaDataStream(mdRawData);
396
397		//parse meta-data file
398	<	Globals* simParams = parseFile(rawMetaDataStream, mdFileName, metaDataBlockStart+1);
398	>	Globals* simParams = parseFile(rawMetaDataStream, mdFileName, mdFileVersion,
399	>	metaDataBlockStart + 1);
400
401		//create the force field
402	<	ForceField * ff = ForceFieldFactory::getInstance()
403	<	->createForceField(simParams->getForceField());
335	<
402	>	ForceField * ff = new ForceField(simParams->getForceField());
403	>
404		if (ff == NULL) {
405		sprintf(painCave.errMsg,
406		"ForceField Factory can not create %s force field\n",
#	Line 365 | Line 433 | namespace oopse {
433		}
434
435		ff->parse(forcefieldFileName);
368	–	ff->setFortranForceOptions();
436		//create SimInfo
437		SimInfo * info = new SimInfo(ff, simParams);
438
#	Line 383 | Line 450 | namespace oopse {
450		//create the molecules
451		createMolecules(info);
452
453	<
453	>	//find the storage layout
454	>
455	>	int storageLayout = computeStorageLayout(info);
456	>
457		//allocate memory for DataStorage(circular reference, need to
458		//break it)
459	<	info->setSnapshotManager(new SimSnapshotManager(info));
459	>	info->setSnapshotManager(new SimSnapshotManager(info, storageLayout));
460
461		//set the global index of atoms, rigidbodies and cutoffgroups
462		//(only need to be set once, the global index will never change
#	Line 395 | Line 465 | namespace oopse {
465		//responsibility to LocalIndexManager.
466		setGlobalIndex(info);
467
468	<	//Although addExcludePairs is called inside SimInfo's addMolecule
468	>	//Although addInteractionPairs is called inside SimInfo's addMolecule
469		//method, at that point atoms don't have the global index yet
470		//(their global index are all initialized to -1).  Therefore we
471	<	//have to call addExcludePairs explicitly here. A way to work
471	>	//have to call addInteractionPairs explicitly here. A way to work
472		//around is that we can determine the beginning global indices of
473		//atoms before they get created.
474		SimInfo::MoleculeIterator mi;
475		Molecule* mol;
476		for (mol= info->beginMolecule(mi); mol != NULL; mol = info->nextMolecule(mi)) {
477	<	info->addExcludePairs(mol);
477	>	info->addInteractionPairs(mol);
478		}
479
480		if (loadInitCoords)
481		loadCoordinates(info, mdFileName);
412	–
482		return info;
483		}
484
#	Line 444 | Line 513 | namespace oopse {
513
514		#ifdef IS_MPI
515		void SimCreator::divideMolecules(SimInfo *info) {
447	–	RealType numerator;
448	–	RealType denominator;
449	–	RealType precast;
450	–	RealType x;
451	–	RealType y;
516		RealType a;
453	–	int old_atoms;
454	–	int add_atoms;
455	–	int new_atoms;
456	–	int nTarget;
457	–	int done;
458	–	int i;
459	–	int j;
460	–	int loops;
461	–	int which_proc;
517		int nProcessors;
518		std::vector<int> atomsPerProc;
519		int nGlobalMols = info->getNGlobalMolecules();
520	<	std::vector<int> molToProcMap(nGlobalMols, -1); // default to an error condition:
520	>	std::vector<int> molToProcMap(nGlobalMols, -1); // default to an
521	>	// error
522	>	// condition:
523
524	<	MPI_Comm_size(MPI_COMM_WORLD, &nProcessors);
524	>	MPI_Comm_size( MPI_COMM_WORLD, &nProcessors);
525
526		if (nProcessors > nGlobalMols) {
527		sprintf(painCave.errMsg,
#	Line 473 | Line 530 | namespace oopse {
530		"\tthe number of molecules.  This will not result in a \n"
531		"\tusable division of atoms for force decomposition.\n"
532		"\tEither try a smaller number of processors, or run the\n"
533	<	"\tsingle-processor version of OOPSE.\n", nProcessors, nGlobalMols);
533	>	"\tsingle-processor version of OpenMD.\n", nProcessors,
534	>	nGlobalMols);
535
536		painCave.isFatal = 1;
537		simError();
538		}
539
482	–	int seedValue;
540		Globals * simParams = info->getSimParams();
541	<	SeqRandNumGen* myRandom; //divide labor does not need Parallel random number generator
541	>	SeqRandNumGen* myRandom; //divide labor does not need Parallel
542	>	//random number generator
543		if (simParams->haveSeed()) {
544	<	seedValue = simParams->getSeed();
544	>	int seedValue = simParams->getSeed();
545		myRandom = new SeqRandNumGen(seedValue);
546		}else {
547		myRandom = new SeqRandNumGen();
#	Line 496 | Line 554 | namespace oopse {
554		atomsPerProc.insert(atomsPerProc.end(), nProcessors, 0);
555
556		if (worldRank == 0) {
557	<	numerator = info->getNGlobalAtoms();
558	<	denominator = nProcessors;
559	<	precast = numerator / denominator;
560	<	nTarget = (int)(precast + 0.5);
557	>	RealType numerator = info->getNGlobalAtoms();
558	>	RealType denominator = nProcessors;
559	>	RealType precast = numerator / denominator;
560	>	int nTarget = (int)(precast + 0.5);
561
562	<	for(i = 0; i < nGlobalMols; i++) {
563	<	done = 0;
564	<	loops = 0;
562	>	for(int i = 0; i < nGlobalMols; i++) {
563	>
564	>	int done = 0;
565	>	int loops = 0;
566
567		while (!done) {
568		loops++;
569
570		// Pick a processor at random
571
572	<	which_proc = (int) (myRandom->rand() * nProcessors);
572	>	int which_proc = (int) (myRandom->rand() * nProcessors);
573
574		//get the molecule stamp first
575		int stampId = info->getMoleculeStampId(i);
576		MoleculeStamp * moleculeStamp = info->getMoleculeStamp(stampId);
577
578		// How many atoms does this processor have so far?
579	<	old_atoms = atomsPerProc[which_proc];
580	<	add_atoms = moleculeStamp->getNAtoms();
581	<	new_atoms = old_atoms + add_atoms;
579	>	int old_atoms = atomsPerProc[which_proc];
580	>	int add_atoms = moleculeStamp->getNAtoms();
581	>	int new_atoms = old_atoms + add_atoms;
582
583		// If we've been through this loop too many times, we need
584		// to just give up and assign the molecule to this processor
585		// and be done with it.
586
587		if (loops > 100) {
588	+
589		sprintf(painCave.errMsg,
590	<	"I've tried 100 times to assign molecule %d to a "
591	<	" processor, but can't find a good spot.\n"
592	<	"I'm assigning it at random to processor %d.\n",
590	>	"There have been 100 attempts to assign molecule %d to an\n"
591	>	"\tunderworked processor, but there's no good place to\n"
592	>	"\tleave it.  OpenMD is assigning it at random to processor %d.\n",
593		i, which_proc);
594	<
594	>
595		painCave.isFatal = 0;
596	+	painCave.severity = OPENMD_INFO;
597		simError();
598
599		molToProcMap[i] = which_proc;
#	Line 561 | Line 622 | namespace oopse {
622		//           Pacc(x) = exp(- a * x)
623		// where a = penalty / (average atoms per molecule)
624
625	<	x = (RealType)(new_atoms - nTarget);
626	<	y = myRandom->rand();
625	>	RealType x = (RealType)(new_atoms - nTarget);
626	>	RealType y = myRandom->rand();
627
628		if (y < exp(- a * x)) {
629		molToProcMap[i] = which_proc;
#	Line 577 | Line 638 | namespace oopse {
638		}
639
640		delete myRandom;
641	<
641	>
642		// Spray out this nonsense to all other processors:
582	–
643		MPI_Bcast(&molToProcMap[0], nGlobalMols, MPI_INT, 0, MPI_COMM_WORLD);
644	+
645		} else {
646
647		// Listen to your marching orders from processor 0:
587	–
648		MPI_Bcast(&molToProcMap[0], nGlobalMols, MPI_INT, 0, MPI_COMM_WORLD);
649	+
650		}
651
652		info->setMolToProcMap(molToProcMap);
653		sprintf(checkPointMsg,
654		"Successfully divided the molecules among the processors.\n");
655	<	MPIcheckPoint();
655	>	errorCheckPoint();
656		}
657
658		#endif
#	Line 608 | Line 669 | namespace oopse {
669		#endif
670
671		stampId = info->getMoleculeStampId(i);
672	<	Molecule * mol = molCreator.createMolecule(info->getForceField(), info->getMoleculeStamp(stampId),
673	<	stampId, i, info->getLocalIndexManager());
672	>	Molecule * mol = molCreator.createMolecule(info->getForceField(),
673	>	info->getMoleculeStamp(stampId),
674	>	stampId, i,
675	>	info->getLocalIndexManager());
676
677		info->addMolecule(mol);
678
#	Line 621 | Line 684 | namespace oopse {
684
685		} //end for(int i=0)
686		}
687	+
688	+	int SimCreator::computeStorageLayout(SimInfo* info) {
689	+
690	+	Globals* simParams = info->getSimParams();
691	+	int nRigidBodies = info->getNGlobalRigidBodies();
692	+	set<AtomType*> atomTypes = info->getSimulatedAtomTypes();
693	+	set<AtomType*>::iterator i;
694	+	bool hasDirectionalAtoms = false;
695	+	bool hasFixedCharge = false;
696	+	bool hasDipoles = false;
697	+	bool hasQuadrupoles = false;
698	+	bool hasPolarizable = false;
699	+	bool hasFluctuatingCharge = false;
700	+	bool hasMetallic = false;
701	+	int storageLayout = 0;
702	+	storageLayout |= DataStorage::dslPosition;
703	+	storageLayout |= DataStorage::dslVelocity;
704	+	storageLayout |= DataStorage::dslForce;
705	+
706	+	for (i = atomTypes.begin(); i != atomTypes.end(); ++i) {
707	+
708	+	DirectionalAdapter da = DirectionalAdapter( (*i) );
709	+	MultipoleAdapter ma = MultipoleAdapter( (*i) );
710	+	EAMAdapter ea = EAMAdapter( (*i) );
711	+	SuttonChenAdapter sca = SuttonChenAdapter( (*i) );
712	+	PolarizableAdapter pa = PolarizableAdapter( (*i) );
713	+	FixedChargeAdapter fca = FixedChargeAdapter( (*i) );
714	+	FluctuatingChargeAdapter fqa = FluctuatingChargeAdapter( (*i) );
715	+
716	+	if (da.isDirectional()){
717	+	hasDirectionalAtoms = true;
718	+	}
719	+	if (ma.isDipole()){
720	+	hasDipoles = true;
721	+	}
722	+	if (ma.isQuadrupole()){
723	+	hasQuadrupoles = true;
724	+	}
725	+	if (ea.isEAM() || sca.isSuttonChen()){
726	+	hasMetallic = true;
727	+	}
728	+	if ( fca.isFixedCharge() ){
729	+	hasFixedCharge = true;
730	+	}
731	+	if ( fqa.isFluctuatingCharge() ){
732	+	hasFluctuatingCharge = true;
733	+	}
734	+	if ( pa.isPolarizable() ){
735	+	hasPolarizable = true;
736	+	}
737	+	}
738
739	+	if (nRigidBodies > 0 || hasDirectionalAtoms) {
740	+	storageLayout |= DataStorage::dslAmat;
741	+	if(storageLayout & DataStorage::dslVelocity) {
742	+	storageLayout |= DataStorage::dslAngularMomentum;
743	+	}
744	+	if (storageLayout & DataStorage::dslForce) {
745	+	storageLayout |= DataStorage::dslTorque;
746	+	}
747	+	}
748	+	if (hasDipoles) {
749	+	storageLayout |= DataStorage::dslDipole;
750	+	}
751	+	if (hasQuadrupoles) {
752	+	storageLayout |= DataStorage::dslQuadrupole;
753	+	}
754	+	if (hasFixedCharge || hasFluctuatingCharge) {
755	+	storageLayout |= DataStorage::dslSkippedCharge;
756	+	}
757	+	if (hasMetallic) {
758	+	storageLayout |= DataStorage::dslDensity;
759	+	storageLayout |= DataStorage::dslFunctional;
760	+	storageLayout |= DataStorage::dslFunctionalDerivative;
761	+	}
762	+	if (hasPolarizable) {
763	+	storageLayout |= DataStorage::dslElectricField;
764	+	}
765	+	if (hasFluctuatingCharge){
766	+	storageLayout |= DataStorage::dslFlucQPosition;
767	+	if(storageLayout & DataStorage::dslVelocity) {
768	+	storageLayout |= DataStorage::dslFlucQVelocity;
769	+	}
770	+	if (storageLayout & DataStorage::dslForce) {
771	+	storageLayout |= DataStorage::dslFlucQForce;
772	+	}
773	+	}
774	+
775	+	// if the user has asked for them, make sure we've got the memory for the
776	+	// objects defined.
777	+
778	+	if (simParams->getOutputParticlePotential()) {
779	+	storageLayout |= DataStorage::dslParticlePot;
780	+	}
781	+
782	+	if (simParams->havePrintHeatFlux()) {
783	+	if (simParams->getPrintHeatFlux()) {
784	+	storageLayout |= DataStorage::dslParticlePot;
785	+	}
786	+	}
787	+
788	+	if (simParams->getOutputElectricField() | simParams->haveElectricField()) {
789	+	storageLayout |= DataStorage::dslElectricField;
790	+	}
791	+
792	+	if (simParams->getOutputSitePotential() ) {
793	+	storageLayout |= DataStorage::dslSitePotential;
794	+	}
795	+
796	+	if (simParams->getOutputFluctuatingCharges()) {
797	+	storageLayout |= DataStorage::dslFlucQPosition;
798	+	storageLayout |= DataStorage::dslFlucQVelocity;
799	+	storageLayout |= DataStorage::dslFlucQForce;
800	+	}
801	+
802	+	info->setStorageLayout(storageLayout);
803	+
804	+	return storageLayout;
805	+	}
806	+
807		void SimCreator::setGlobalIndex(SimInfo *info) {
808		SimInfo::MoleculeIterator mi;
809		Molecule::AtomIterator ai;
810		Molecule::RigidBodyIterator ri;
811		Molecule::CutoffGroupIterator ci;
812	+	Molecule::BondIterator boi;
813	+	Molecule::BendIterator bei;
814	+	Molecule::TorsionIterator ti;
815	+	Molecule::InversionIterator ii;
816		Molecule::IntegrableObjectIterator  ioi;
817	<	Molecule * mol;
818	<	Atom * atom;
819	<	RigidBody * rb;
820	<	CutoffGroup * cg;
817	>	Molecule* mol;
818	>	Atom* atom;
819	>	RigidBody* rb;
820	>	CutoffGroup* cg;
821	>	Bond* bond;
822	>	Bend* bend;
823	>	Torsion* torsion;
824	>	Inversion* inversion;
825		int beginAtomIndex;
826		int beginRigidBodyIndex;
827		int beginCutoffGroupIndex;
828	+	int beginBondIndex;
829	+	int beginBendIndex;
830	+	int beginTorsionIndex;
831	+	int beginInversionIndex;
832		int nGlobalAtoms = info->getNGlobalAtoms();
833	<
640	<	/**@todo fixme */
641	<	#ifndef IS_MPI
833	>	int nGlobalRigidBodies = info->getNGlobalRigidBodies();
834
835		beginAtomIndex = 0;
836	<	beginRigidBodyIndex = 0;
836	>	// The rigid body indices begin immediately after the atom indices:
837	>	beginRigidBodyIndex = info->getNGlobalAtoms();
838		beginCutoffGroupIndex = 0;
839	<
840	<	#else
841	<
842	<	int nproc;
843	<	int myNode;
844	<
652	<	myNode = worldRank;
653	<	MPI_Comm_size(MPI_COMM_WORLD, &nproc);
654	<
655	<	std::vector < int > tmpAtomsInProc(nproc, 0);
656	<	std::vector < int > tmpRigidBodiesInProc(nproc, 0);
657	<	std::vector < int > tmpCutoffGroupsInProc(nproc, 0);
658	<	std::vector < int > NumAtomsInProc(nproc, 0);
659	<	std::vector < int > NumRigidBodiesInProc(nproc, 0);
660	<	std::vector < int > NumCutoffGroupsInProc(nproc, 0);
661	<
662	<	tmpAtomsInProc[myNode] = info->getNAtoms();
663	<	tmpRigidBodiesInProc[myNode] = info->getNRigidBodies();
664	<	tmpCutoffGroupsInProc[myNode] = info->getNCutoffGroups();
665	<
666	<	//do MPI_ALLREDUCE to exchange the total number of atoms, rigidbodies and cutoff groups
667	<	MPI_Allreduce(&tmpAtomsInProc[0], &NumAtomsInProc[0], nproc, MPI_INT,
668	<	MPI_SUM, MPI_COMM_WORLD);
669	<	MPI_Allreduce(&tmpRigidBodiesInProc[0], &NumRigidBodiesInProc[0], nproc,
670	<	MPI_INT, MPI_SUM, MPI_COMM_WORLD);
671	<	MPI_Allreduce(&tmpCutoffGroupsInProc[0], &NumCutoffGroupsInProc[0], nproc,
672	<	MPI_INT, MPI_SUM, MPI_COMM_WORLD);
673	<
674	<	beginAtomIndex = 0;
675	<	beginRigidBodyIndex = 0;
676	<	beginCutoffGroupIndex = 0;
677	<
678	<	for(int i = 0; i < myNode; i++) {
679	<	beginAtomIndex += NumAtomsInProc[i];
680	<	beginRigidBodyIndex += NumRigidBodiesInProc[i];
681	<	beginCutoffGroupIndex += NumCutoffGroupsInProc[i];
682	<	}
683	<
684	<	#endif
685	<
686	<	//rigidbody's index begins right after atom's
687	<	beginRigidBodyIndex += info->getNGlobalAtoms();
688	<
689	<	for(mol = info->beginMolecule(mi); mol != NULL;
690	<	mol = info->nextMolecule(mi)) {
839	>	beginBondIndex = 0;
840	>	beginBendIndex = 0;
841	>	beginTorsionIndex = 0;
842	>	beginInversionIndex = 0;
843	>
844	>	for(int i = 0; i < info->getNGlobalMolecules(); i++) {
845
846	<	//local index(index in DataStorge) of atom is important
847	<	for(atom = mol->beginAtom(ai); atom != NULL; atom = mol->nextAtom(ai)) {
848	<	atom->setGlobalIndex(beginAtomIndex++);
846	>	#ifdef IS_MPI
847	>	if (info->getMolToProc(i) == worldRank) {
848	>	#endif
849	>	// stuff to do if I own this molecule
850	>	mol = info->getMoleculeByGlobalIndex(i);
851	>
852	>	// The local index(index in DataStorge) of the atom is important:
853	>	for(atom = mol->beginAtom(ai); atom != NULL;
854	>	atom = mol->nextAtom(ai)) {
855	>	atom->setGlobalIndex(beginAtomIndex++);
856	>	}
857	>
858	>	for(rb = mol->beginRigidBody(ri); rb != NULL;
859	>	rb = mol->nextRigidBody(ri)) {
860	>	rb->setGlobalIndex(beginRigidBodyIndex++);
861	>	}
862	>
863	>	// The local index of other objects only depends on the order
864	>	// of traversal:
865	>	for(cg = mol->beginCutoffGroup(ci); cg != NULL;
866	>	cg = mol->nextCutoffGroup(ci)) {
867	>	cg->setGlobalIndex(beginCutoffGroupIndex++);
868	>	}
869	>	for(bond = mol->beginBond(boi); bond != NULL;
870	>	bond = mol->nextBond(boi)) {
871	>	bond->setGlobalIndex(beginBondIndex++);
872	>	}
873	>	for(bend = mol->beginBend(bei); bend != NULL;
874	>	bend = mol->nextBend(bei)) {
875	>	bend->setGlobalIndex(beginBendIndex++);
876	>	}
877	>	for(torsion = mol->beginTorsion(ti); torsion != NULL;
878	>	torsion = mol->nextTorsion(ti)) {
879	>	torsion->setGlobalIndex(beginTorsionIndex++);
880	>	}
881	>	for(inversion = mol->beginInversion(ii); inversion != NULL;
882	>	inversion = mol->nextInversion(ii)) {
883	>	inversion->setGlobalIndex(beginInversionIndex++);
884	>	}
885	>
886	>	#ifdef IS_MPI
887	>	}  else {
888	>
889	>	// stuff to do if I don't own this molecule
890	>
891	>	int stampId = info->getMoleculeStampId(i);
892	>	MoleculeStamp* stamp = info->getMoleculeStamp(stampId);
893	>
894	>	beginAtomIndex += stamp->getNAtoms();
895	>	beginRigidBodyIndex += stamp->getNRigidBodies();
896	>	beginCutoffGroupIndex += stamp->getNCutoffGroups() + stamp->getNFreeAtoms();
897	>	beginBondIndex += stamp->getNBonds();
898	>	beginBendIndex += stamp->getNBends();
899	>	beginTorsionIndex += stamp->getNTorsions();
900	>	beginInversionIndex += stamp->getNInversions();
901		}
902	<
903	<	for(rb = mol->beginRigidBody(ri); rb != NULL;
904	<	rb = mol->nextRigidBody(ri)) {
905	<	rb->setGlobalIndex(beginRigidBodyIndex++);
700	<	}
701	<
702	<	//local index of cutoff group is trivial, it only depends on the order of travesing
703	<	for(cg = mol->beginCutoffGroup(ci); cg != NULL;
704	<	cg = mol->nextCutoffGroup(ci)) {
705	<	cg->setGlobalIndex(beginCutoffGroupIndex++);
706	<	}
707	<	}
708	<
902	>	#endif
903	>
904	>	} //end for(int i=0)
905	>
906		//fill globalGroupMembership
907		std::vector<int> globalGroupMembership(info->getNGlobalAtoms(), 0);
908	<	for(mol = info->beginMolecule(mi); mol != NULL; mol = info->nextMolecule(mi)) {
909	<	for (cg = mol->beginCutoffGroup(ci); cg != NULL; cg = mol->nextCutoffGroup(ci)) {
910	<
908	>	for(mol = info->beginMolecule(mi); mol != NULL;
909	>	mol = info->nextMolecule(mi)) {
910	>	for (cg = mol->beginCutoffGroup(ci); cg != NULL;
911	>	cg = mol->nextCutoffGroup(ci)) {
912		for(atom = cg->beginAtom(ai); atom != NULL; atom = cg->nextAtom(ai)) {
913		globalGroupMembership[atom->getGlobalIndex()] = cg->getGlobalIndex();
914		}
915
916		}
917		}
918	<
918	>
919		#ifdef IS_MPI
920		// Since the globalGroupMembership has been zero filled and we've only
921		// poked values into the atoms we know, we can do an Allreduce
922		// to get the full globalGroupMembership array (We think).
923		// This would be prettier if we could use MPI_IN_PLACE like the MPI-2
924		// docs said we could.
925	<	std::vector<int> tmpGroupMembership(nGlobalAtoms, 0);
926	<	MPI_Allreduce(&globalGroupMembership[0], &tmpGroupMembership[0], nGlobalAtoms,
925	>	std::vector<int> tmpGroupMembership(info->getNGlobalAtoms(), 0);
926	>	MPI_Allreduce(&globalGroupMembership[0],
927	>	&tmpGroupMembership[0], nGlobalAtoms,
928		MPI_INT, MPI_SUM, MPI_COMM_WORLD);
929	+
930		info->setGlobalGroupMembership(tmpGroupMembership);
931		#else
932		info->setGlobalGroupMembership(globalGroupMembership);
933		#endif
934
935		//fill molMembership
936	<	std::vector<int> globalMolMembership(info->getNGlobalAtoms(), 0);
936	>	std::vector<int> globalMolMembership(info->getNGlobalAtoms() +
937	>	info->getNGlobalRigidBodies(), 0);
938
939	<	for(mol = info->beginMolecule(mi); mol != NULL; mol = info->nextMolecule(mi)) {
940	<
939	>	for(mol = info->beginMolecule(mi); mol != NULL;
940	>	mol = info->nextMolecule(mi)) {
941		for(atom = mol->beginAtom(ai); atom != NULL; atom = mol->nextAtom(ai)) {
942		globalMolMembership[atom->getGlobalIndex()] = mol->getGlobalIndex();
943		}
944	+	for (rb = mol->beginRigidBody(ri); rb != NULL;
945	+	rb = mol->nextRigidBody(ri)) {
946	+	globalMolMembership[rb->getGlobalIndex()] = mol->getGlobalIndex();
947	+	}
948		}
949
950		#ifdef IS_MPI
951	<	std::vector<int> tmpMolMembership(nGlobalAtoms, 0);
952	<
953	<	MPI_Allreduce(&globalMolMembership[0], &tmpMolMembership[0], nGlobalAtoms,
951	>	std::vector<int> tmpMolMembership(info->getNGlobalAtoms() +
952	>	info->getNGlobalRigidBodies(), 0);
953	>	MPI_Allreduce(&globalMolMembership[0], &tmpMolMembership[0],
954	>	nGlobalAtoms + nGlobalRigidBodies,
955		MPI_INT, MPI_SUM, MPI_COMM_WORLD);
956
957		info->setGlobalMolMembership(tmpMolMembership);
#	Line 757 | Line 963 | namespace oopse {
963		// here the molecules are listed by their global indices.
964
965		std::vector<int> nIOPerMol(info->getNGlobalMolecules(), 0);
966	<	for (mol = info->beginMolecule(mi); mol != NULL; mol = info->nextMolecule(mi)) {
966	>	for (mol = info->beginMolecule(mi); mol != NULL;
967	>	mol = info->nextMolecule(mi)) {
968		nIOPerMol[mol->getGlobalIndex()] = mol->getNIntegrableObjects();
969		}
970
971		#ifdef IS_MPI
972		std::vector<int> numIntegrableObjectsPerMol(info->getNGlobalMolecules(), 0);
973		MPI_Allreduce(&nIOPerMol[0], &numIntegrableObjectsPerMol[0],
974	<	info->getNGlobalMolecules(), MPI_INT, MPI_SUM, MPI_COMM_WORLD);
974	>	info->getNGlobalMolecules(), MPI_INT, MPI_SUM, MPI_COMM_WORLD);
975		#else
976		std::vector<int> numIntegrableObjectsPerMol = nIOPerMol;
977		#endif
978
979	<	std::vector<int> startingIOIndexForMol(info->getNGlobalMolecules());
980	<
981	<	int startingIndex = 0;
982	<	for (int i = 0; i < info->getNGlobalMolecules(); i++) {
983	<	startingIOIndexForMol[i] = startingIndex;
984	<	startingIndex += numIntegrableObjectsPerMol[i];
985	<	}
986	<
987	<	std::vector<StuntDouble*> IOIndexToIntegrableObject(info->getNGlobalIntegrableObjects(), (StuntDouble*)NULL);
988	<	for (mol = info->beginMolecule(mi); mol != NULL; mol = info->nextMolecule(mi)) {
979	>	std::vector<int> startingIOIndexForMol(info->getNGlobalMolecules());
980	>
981	>	int startingIndex = 0;
982	>	for (int i = 0; i < info->getNGlobalMolecules(); i++) {
983	>	startingIOIndexForMol[i] = startingIndex;
984	>	startingIndex += numIntegrableObjectsPerMol[i];
985	>	}
986	>
987	>	std::vector<StuntDouble*> IOIndexToIntegrableObject(info->getNGlobalIntegrableObjects(), (StuntDouble*)NULL);
988	>	for (mol = info->beginMolecule(mi); mol != NULL;
989	>	mol = info->nextMolecule(mi)) {
990		int myGlobalIndex = mol->getGlobalIndex();
991		int globalIO = startingIOIndexForMol[myGlobalIndex];
992	<	for (StuntDouble* integrableObject = mol->beginIntegrableObject(ioi); integrableObject != NULL;
993	<	integrableObject = mol->nextIntegrableObject(ioi)) {
994	<	integrableObject->setGlobalIntegrableObjectIndex(globalIO);
995	<	IOIndexToIntegrableObject[globalIO] = integrableObject;
996	<	globalIO++;
992	>	for (StuntDouble* sd = mol->beginIntegrableObject(ioi); sd != NULL;
993	>	sd = mol->nextIntegrableObject(ioi)) {
994	>	sd->setGlobalIntegrableObjectIndex(globalIO);
995	>	IOIndexToIntegrableObject[globalIO] = sd;
996	>	globalIO++;
997		}
998		}
999	<
1000	<	info->setIOIndexToIntegrableObject(IOIndexToIntegrableObject);
1001	<
999	>
1000	>	info->setIOIndexToIntegrableObject(IOIndexToIntegrableObject);
1001	>
1002		}
1003
1004		void SimCreator::loadCoordinates(SimInfo* info, const std::string& mdFileName) {
797	–	Globals* simParams;
798	–	simParams = info->getSimParams();
1005
800	–
1006		DumpReader reader(info, mdFileName);
1007		int nframes = reader.getNFrames();
1008
#	Line 811 | Line 1016 | int startingIndex = 0;
1016		painCave.isFatal = 1;
1017		simError();
1018		}
814	–
1019		//copy the current snapshot to previous snapshot
1020		info->getSnapshotManager()->advance();
1021		}
1022
1023	<	} //end namespace oopse
1023	>	} //end namespace OpenMD
1024
1025

Comparing trunk/src/brains/SimCreator.cpp (property svn:keywords):
Revision 1025 by gezelter, Wed Aug 30 20:33:44 2006 UTC vs.
Revision 1993 by gezelter, Tue Apr 29 17:32:31 2014 UTC

#	Line 0 | Line 1
1	+	Author Id Revision Date

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines