ViewVC Help

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

root/OpenMD/branches/development/src/rnemd/RNEMD.cpp

Comparing:
trunk/src/integrators/RNEMD.cpp (file contents), Revision 1350 by gezelter, Thu May 21 18:56:45 2009 UTC vs.
branches/development/src/rnemd/RNEMD.cpp (file contents), Revision 1830 by gezelter, Wed Jan 9 22:02:30 2013 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, 24107 (2008).
39	+	* [4]  Vardeman & Gezelter, in progress (2009).
40		*/
41
42	<	#include "integrators/RNEMD.hpp"
42	>	#include <cmath>
43	>	#include "rnemd/RNEMD.hpp"
44		#include "math/Vector3.hpp"
45	+	#include "math/Vector.hpp"
46		#include "math/SquareMatrix3.hpp"
47	+	#include "math/Polynomial.hpp"
48		#include "primitives/Molecule.hpp"
49		#include "primitives/StuntDouble.hpp"
50	<	#include "utils/OOPSEConstant.hpp"
50	>	#include "utils/PhysicalConstants.hpp"
51		#include "utils/Tuple.hpp"
52	+	#ifdef IS_MPI
53	+	#include <mpi.h>
54	+	#endif
55
56	<	#ifndef IS_MPI
57	<	#include "math/SeqRandNumGen.hpp"
58	<	#else
53	<	#include "math/ParallelRandNumGen.hpp"
56	>	#ifdef _MSC_VER
57	>	#define isnan(x) _isnan((x))
58	>	#define isinf(x) (!_finite(x) && !_isnan(x))
59		#endif
60
61		#define HONKING_LARGE_VALUE 1.0e10
62
63	<	namespace oopse {
63	>	using namespace std;
64	>	namespace OpenMD {
65
66	<	RNEMD::RNEMD(SimInfo* info) : info_(info), evaluator_(info), seleMan_(info), usePeriodicBoundaryConditions_(info->getSimParams()->getUsePeriodicBoundaryConditions()) {
67	<
68	<	int seedValue;
66	>	RNEMD::RNEMD(SimInfo* info) : info_(info), evaluator_(info), seleMan_(info),
67	>	usePeriodicBoundaryConditions_(info->getSimParams()->getUsePeriodicBoundaryConditions()) {
68	>
69	>	trialCount_ = 0;
70	>	failTrialCount_ = 0;
71	>	failRootCount_ = 0;
72	>
73		Globals * simParams = info->getSimParams();
74	+	RNEMDParameters* rnemdParams = simParams->getRNEMDParameters();
75
76	<	stringToEnumMap_["Kinetic"] = rnemdKinetic;
77	<	stringToEnumMap_["Px"] = rnemdPx;
67	<	stringToEnumMap_["Py"] = rnemdPy;
68	<	stringToEnumMap_["Pz"] = rnemdPz;
69	<	stringToEnumMap_["Unknown"] = rnemdUnknown;
76	>	doRNEMD_ = rnemdParams->getUseRNEMD();
77	>	if (!doRNEMD_) return;
78
79	<	rnemdObjectSelection_ = simParams->getRNEMD_objectSelection();
79	>	stringToMethod_["Swap"]  = rnemdSwap;
80	>	stringToMethod_["NIVS"]  = rnemdNIVS;
81	>	stringToMethod_["VSS"]   = rnemdVSS;
82	>
83	>	stringToFluxType_["KE"]  = rnemdKE;
84	>	stringToFluxType_["Px"]  = rnemdPx;
85	>	stringToFluxType_["Py"]  = rnemdPy;
86	>	stringToFluxType_["Pz"]  = rnemdPz;
87	>	stringToFluxType_["Pvector"]  = rnemdPvector;
88	>	stringToFluxType_["KE+Px"]  = rnemdKePx;
89	>	stringToFluxType_["KE+Py"]  = rnemdKePy;
90	>	stringToFluxType_["KE+Pvector"]  = rnemdKePvector;
91	>
92	>	runTime_ = simParams->getRunTime();
93	>	statusTime_ = simParams->getStatusTime();
94	>
95	>	rnemdObjectSelection_ = rnemdParams->getObjectSelection();
96		evaluator_.loadScriptString(rnemdObjectSelection_);
97		seleMan_.setSelectionSet(evaluator_.evaluate());
98
99	+	const string methStr = rnemdParams->getMethod();
100	+	bool hasFluxType = rnemdParams->haveFluxType();
101
102	+	string fluxStr;
103	+	if (hasFluxType) {
104	+	fluxStr = rnemdParams->getFluxType();
105	+	} else {
106	+	sprintf(painCave.errMsg,
107	+	"RNEMD: No fluxType was set in the md file.  This parameter,\n"
108	+	"\twhich must be one of the following values:\n"
109	+	"\tKE, Px, Py, Pz, Pvector, KE+Px, KE+Py, KE+Pvector\n"
110	+	"\tmust be set to use RNEMD\n");
111	+	painCave.isFatal = 1;
112	+	painCave.severity = OPENMD_ERROR;
113	+	simError();
114	+	}
115	+
116	+	bool hasKineticFlux = rnemdParams->haveKineticFlux();
117	+	bool hasMomentumFlux = rnemdParams->haveMomentumFlux();
118	+	bool hasMomentumFluxVector = rnemdParams->haveMomentumFluxVector();
119	+	bool hasSlabWidth = rnemdParams->haveSlabWidth();
120	+	bool hasSlabACenter = rnemdParams->haveSlabACenter();
121	+	bool hasSlabBCenter = rnemdParams->haveSlabBCenter();
122	+	bool hasOutputFileName = rnemdParams->haveOutputFileName();
123	+	bool hasOutputFields = rnemdParams->haveOutputFields();
124	+
125	+	map<string, RNEMDMethod>::iterator i;
126	+	i = stringToMethod_.find(methStr);
127	+	if (i != stringToMethod_.end())
128	+	rnemdMethod_ = i->second;
129	+	else {
130	+	sprintf(painCave.errMsg,
131	+	"RNEMD: The current method,\n"
132	+	"\t\t%s is not one of the recognized\n"
133	+	"\texchange methods: Swap, NIVS, or VSS\n",
134	+	methStr.c_str());
135	+	painCave.isFatal = 1;
136	+	painCave.severity = OPENMD_ERROR;
137	+	simError();
138	+	}
139	+
140	+	map<string, RNEMDFluxType>::iterator j;
141	+	j = stringToFluxType_.find(fluxStr);
142	+	if (j != stringToFluxType_.end())
143	+	rnemdFluxType_ = j->second;
144	+	else {
145	+	sprintf(painCave.errMsg,
146	+	"RNEMD: The current fluxType,\n"
147	+	"\t\t%s\n"
148	+	"\tis not one of the recognized flux types.\n",
149	+	fluxStr.c_str());
150	+	painCave.isFatal = 1;
151	+	painCave.severity = OPENMD_ERROR;
152	+	simError();
153	+	}
154	+
155	+	bool methodFluxMismatch = false;
156	+	bool hasCorrectFlux = false;
157	+	switch(rnemdMethod_) {
158	+	case rnemdSwap:
159	+	switch (rnemdFluxType_) {
160	+	case rnemdKE:
161	+	hasCorrectFlux = hasKineticFlux;
162	+	break;
163	+	case rnemdPx:
164	+	case rnemdPy:
165	+	case rnemdPz:
166	+	hasCorrectFlux = hasMomentumFlux;
167	+	break;
168	+	default :
169	+	methodFluxMismatch = true;
170	+	break;
171	+	}
172	+	break;
173	+	case rnemdNIVS:
174	+	switch (rnemdFluxType_) {
175	+	case rnemdKE:
176	+	case rnemdRotKE:
177	+	case rnemdFullKE:
178	+	hasCorrectFlux = hasKineticFlux;
179	+	break;
180	+	case rnemdPx:
181	+	case rnemdPy:
182	+	case rnemdPz:
183	+	hasCorrectFlux = hasMomentumFlux;
184	+	break;
185	+	case rnemdKePx:
186	+	case rnemdKePy:
187	+	hasCorrectFlux = hasMomentumFlux && hasKineticFlux;
188	+	break;
189	+	default:
190	+	methodFluxMismatch = true;
191	+	break;
192	+	}
193	+	break;
194	+	case rnemdVSS:
195	+	switch (rnemdFluxType_) {
196	+	case rnemdKE:
197	+	case rnemdRotKE:
198	+	case rnemdFullKE:
199	+	hasCorrectFlux = hasKineticFlux;
200	+	break;
201	+	case rnemdPx:
202	+	case rnemdPy:
203	+	case rnemdPz:
204	+	hasCorrectFlux = hasMomentumFlux;
205	+	break;
206	+	case rnemdPvector:
207	+	hasCorrectFlux = hasMomentumFluxVector;
208	+	break;
209	+	case rnemdKePx:
210	+	case rnemdKePy:
211	+	hasCorrectFlux = hasMomentumFlux && hasKineticFlux;
212	+	break;
213	+	case rnemdKePvector:
214	+	hasCorrectFlux = hasMomentumFluxVector && hasKineticFlux;
215	+	break;
216	+	default:
217	+	methodFluxMismatch = true;
218	+	break;
219	+	}
220	+	default:
221	+	break;
222	+	}
223	+
224	+	if (methodFluxMismatch) {
225	+	sprintf(painCave.errMsg,
226	+	"RNEMD: The current method,\n"
227	+	"\t\t%s\n"
228	+	"\tcannot be used with the current flux type, %s\n",
229	+	methStr.c_str(), fluxStr.c_str());
230	+	painCave.isFatal = 1;
231	+	painCave.severity = OPENMD_ERROR;
232	+	simError();
233	+	}
234	+	if (!hasCorrectFlux) {
235	+	sprintf(painCave.errMsg,
236	+	"RNEMD: The current method, %s, and flux type, %s,\n"
237	+	"\tdid not have the correct flux value specified. Options\n"
238	+	"\tinclude: kineticFlux, momentumFlux, and momentumFluxVector\n",
239	+	methStr.c_str(), fluxStr.c_str());
240	+	painCave.isFatal = 1;
241	+	painCave.severity = OPENMD_ERROR;
242	+	simError();
243	+	}
244	+
245	+	if (hasKineticFlux) {
246	+	// convert the kcal / mol / Angstroms^2 / fs values in the md file
247	+	// into  amu / fs^3:
248	+	kineticFlux_ = rnemdParams->getKineticFlux()
249	+	* PhysicalConstants::energyConvert;
250	+	} else {
251	+	kineticFlux_ = 0.0;
252	+	}
253	+	if (hasMomentumFluxVector) {
254	+	momentumFluxVector_ = rnemdParams->getMomentumFluxVector();
255	+	} else {
256	+	momentumFluxVector_ = V3Zero;
257	+	if (hasMomentumFlux) {
258	+	RealType momentumFlux = rnemdParams->getMomentumFlux();
259	+	switch (rnemdFluxType_) {
260	+	case rnemdPx:
261	+	momentumFluxVector_.x() = momentumFlux;
262	+	break;
263	+	case rnemdPy:
264	+	momentumFluxVector_.y() = momentumFlux;
265	+	break;
266	+	case rnemdPz:
267	+	momentumFluxVector_.z() = momentumFlux;
268	+	break;
269	+	case rnemdKePx:
270	+	momentumFluxVector_.x() = momentumFlux;
271	+	break;
272	+	case rnemdKePy:
273	+	momentumFluxVector_.y() = momentumFlux;
274	+	break;
275	+	default:
276	+	break;
277	+	}
278	+	}
279	+	}
280	+
281		// do some sanity checking
282
283		int selectionCount = seleMan_.getSelectionCount();
284	+
285		int nIntegrable = info->getNGlobalIntegrableObjects();
286
287		if (selectionCount > nIntegrable) {
288		sprintf(painCave.errMsg,
289	<	"RNEMD warning: The current RNEMD_objectSelection,\n"
289	>	"RNEMD: The current objectSelection,\n"
290		"\t\t%s\n"
291		"\thas resulted in %d selected objects.  However,\n"
292		"\tthe total number of integrable objects in the system\n"
#	Line 90 | Line 296 | namespace oopse {
296		rnemdObjectSelection_.c_str(),
297		selectionCount, nIntegrable);
298		painCave.isFatal = 0;
299	+	painCave.severity = OPENMD_WARNING;
300		simError();
301	+	}
302
303	+	areaAccumulator_ = new Accumulator();
304	+
305	+	nBins_ = rnemdParams->getOutputBins();
306	+
307	+	data_.resize(RNEMD::ENDINDEX);
308	+	OutputData z;
309	+	z.units =  "Angstroms";
310	+	z.title =  "Z";
311	+	z.dataType = "RealType";
312	+	z.accumulator.reserve(nBins_);
313	+	for (int i = 0; i < nBins_; i++)
314	+	z.accumulator.push_back( new Accumulator() );
315	+	data_[Z] = z;
316	+	outputMap_["Z"] =  Z;
317	+
318	+	OutputData temperature;
319	+	temperature.units =  "K";
320	+	temperature.title =  "Temperature";
321	+	temperature.dataType = "RealType";
322	+	temperature.accumulator.reserve(nBins_);
323	+	for (int i = 0; i < nBins_; i++)
324	+	temperature.accumulator.push_back( new Accumulator() );
325	+	data_[TEMPERATURE] = temperature;
326	+	outputMap_["TEMPERATURE"] =  TEMPERATURE;
327	+
328	+	OutputData velocity;
329	+	velocity.units = "angstroms/fs";
330	+	velocity.title =  "Velocity";
331	+	velocity.dataType = "Vector3d";
332	+	velocity.accumulator.reserve(nBins_);
333	+	for (int i = 0; i < nBins_; i++)
334	+	velocity.accumulator.push_back( new VectorAccumulator() );
335	+	data_[VELOCITY] = velocity;
336	+	outputMap_["VELOCITY"] = VELOCITY;
337	+
338	+	OutputData density;
339	+	density.units =  "g cm^-3";
340	+	density.title =  "Density";
341	+	density.dataType = "RealType";
342	+	density.accumulator.reserve(nBins_);
343	+	for (int i = 0; i < nBins_; i++)
344	+	density.accumulator.push_back( new Accumulator() );
345	+	data_[DENSITY] = density;
346	+	outputMap_["DENSITY"] =  DENSITY;
347	+
348	+	if (hasOutputFields) {
349	+	parseOutputFileFormat(rnemdParams->getOutputFields());
350	+	} else {
351	+	outputMask_.set(Z);
352	+	switch (rnemdFluxType_) {
353	+	case rnemdKE:
354	+	case rnemdRotKE:
355	+	case rnemdFullKE:
356	+	outputMask_.set(TEMPERATURE);
357	+	break;
358	+	case rnemdPx:
359	+	case rnemdPy:
360	+	outputMask_.set(VELOCITY);
361	+	break;
362	+	case rnemdPz:
363	+	case rnemdPvector:
364	+	outputMask_.set(VELOCITY);
365	+	outputMask_.set(DENSITY);
366	+	break;
367	+	case rnemdKePx:
368	+	case rnemdKePy:
369	+	outputMask_.set(TEMPERATURE);
370	+	outputMask_.set(VELOCITY);
371	+	break;
372	+	case rnemdKePvector:
373	+	outputMask_.set(TEMPERATURE);
374	+	outputMask_.set(VELOCITY);
375	+	outputMask_.set(DENSITY);
376	+	break;
377	+	default:
378	+	break;
379	+	}
380		}
381	<
382	<	const std::string st = simParams->getRNEMD_swapType();
381	>
382	>	if (hasOutputFileName) {
383	>	rnemdFileName_ = rnemdParams->getOutputFileName();
384	>	} else {
385	>	rnemdFileName_ = getPrefix(info->getFinalConfigFileName()) + ".rnemd";
386	>	}
387
388	<	std::map<std::string, RNEMDTypeEnum>::iterator i;
100	<	i = stringToEnumMap_.find(st);
101	<	rnemdType_  = (i == stringToEnumMap_.end()) ? RNEMD::rnemdUnknown : i->second;
388	>	exchangeTime_ = rnemdParams->getExchangeTime();
389
390	<	set_RNEMD_swapTime(simParams->getRNEMD_swapTime());
391	<	set_RNEMD_nBins(simParams->getRNEMD_nBins());
392	<	exchangeSum_ = 0.0;
390	>	Snapshot* currentSnap_ = info->getSnapshotManager()->getCurrentSnapshot();
391	>	Mat3x3d hmat = currentSnap_->getHmat();
392	>
393	>	// Target exchange quantities (in each exchange) =  2 Lx Ly dt flux
394	>	// Lx, Ly = box dimensions in x & y
395	>	// dt = exchange time interval
396	>	// flux = target flux
397
398	<	#ifndef IS_MPI
399	<	if (simParams->haveSeed()) {
400	<	seedValue = simParams->getSeed();
401	<	randNumGen_ = new SeqRandNumGen(seedValue);
402	<	}else {
403	<	randNumGen_ = new SeqRandNumGen();
404	<	}
405	<	#else
406	<	if (simParams->haveSeed()) {
407	<	seedValue = simParams->getSeed();
408	<	randNumGen_ = new ParallelRandNumGen(seedValue);
409	<	}else {
410	<	randNumGen_ = new ParallelRandNumGen();
120	<	}
121	<	#endif
122	<	}
398	>	RealType area = currentSnap_->getXYarea();
399	>	kineticTarget_ = 2.0 * kineticFlux_ * exchangeTime_ * area;
400	>	momentumTarget_ = 2.0 * momentumFluxVector_ * exchangeTime_ * area;
401	>
402	>	// total exchange sums are zeroed out at the beginning:
403	>
404	>	kineticExchange_ = 0.0;
405	>	momentumExchange_ = V3Zero;
406	>
407	>	if (hasSlabWidth)
408	>	slabWidth_ = rnemdParams->getSlabWidth();
409	>	else
410	>	slabWidth_ = hmat(2,2) / 10.0;
411
412	+	if (hasSlabACenter)
413	+	slabACenter_ = rnemdParams->getSlabACenter();
414	+	else
415	+	slabACenter_ = 0.0;
416	+
417	+	if (hasSlabBCenter)
418	+	slabBCenter_ = rnemdParams->getSlabBCenter();
419	+	else
420	+	slabBCenter_ = hmat(2,2) / 2.0;
421	+
422	+	}
423	+
424		RNEMD::~RNEMD() {
425	<	delete randNumGen_;
425	>	if (!doRNEMD_) return;
426	>	#ifdef IS_MPI
427	>	if (worldRank == 0) {
428	>	#endif
429	>
430	>	writeOutputFile();
431	>
432	>	rnemdFile_.close();
433	>
434	>	#ifdef IS_MPI
435	>	}
436	>	#endif
437		}
438	+
439	+	bool RNEMD::inSlabA(Vector3d pos) {
440	+	return (abs(pos.z() - slabACenter_) < 0.5*slabWidth_);
441	+	}
442	+	bool RNEMD::inSlabB(Vector3d pos) {
443	+	return (abs(pos.z() - slabBCenter_) < 0.5*slabWidth_);
444	+	}
445
446		void RNEMD::doSwap() {
447	<	int midBin = nBins_ / 2;
130	<
447	>	if (!doRNEMD_) return;
448		Snapshot* currentSnap_ = info_->getSnapshotManager()->getCurrentSnapshot();
449		Mat3x3d hmat = currentSnap_->getHmat();
450
#	Line 135 | Line 452 | namespace oopse {
452
453		int selei;
454		StuntDouble* sd;
138	–	int idx;
455
456		RealType min_val;
457		bool min_found = false;
#	Line 148 | Line 464 | namespace oopse {
464		for (sd = seleMan_.beginSelected(selei); sd != NULL;
465		sd = seleMan_.nextSelected(selei)) {
466
151	–	idx = sd->getLocalIndex();
152	–
467		Vector3d pos = sd->getPos();
468
469		// wrap the stuntdouble's position back into the box:
470
471		if (usePeriodicBoundaryConditions_)
472		currentSnap_->wrapVector(pos);
473	+	bool inA = inSlabA(pos);
474	+	bool inB = inSlabB(pos);
475
476	<	// which bin is this stuntdouble in?
161	<	// wrapped positions are in the range [-0.5*hmat(2,2), +0.5*hmat(2,2)]
162	<
163	<	int binNo = int(nBins_ * (pos.z() / hmat(2,2) + 0.5)) % nBins_;
164	<
165	<
166	<	// if we're in bin 0 or the middleBin
167	<	if (binNo == 0 || binNo == midBin) {
476	>	if (inA || inB) {
477
478		RealType mass = sd->getMass();
479		Vector3d vel = sd->getVel();
480		RealType value;
481	<
482	<	switch(rnemdType_) {
483	<	case rnemdKinetic :
481	>
482	>	switch(rnemdFluxType_) {
483	>	case rnemdKE :
484
485	<	value = mass * (vel[0]*vel[0] + vel[1]*vel[1] +
486	<	vel[2]*vel[2]);
487	<	if (sd->isDirectional()) {
485	>	value = mass * vel.lengthSquare();
486	>
487	>	if (sd->isDirectional()) {
488		Vector3d angMom = sd->getJ();
489		Mat3x3d I = sd->getI();
490
491		if (sd->isLinear()) {
492	<	int i = sd->linearAxis();
493	<	int j = (i + 1) % 3;
494	<	int k = (i + 2) % 3;
495	<	value += angMom[j] * angMom[j] / I(j, j) +
496	<	angMom[k] * angMom[k] / I(k, k);
492	>	int i = sd->linearAxis();
493	>	int j = (i + 1) % 3;
494	>	int k = (i + 2) % 3;
495	>	value += angMom[j] * angMom[j] / I(j, j) +
496	>	angMom[k] * angMom[k] / I(k, k);
497		} else {
498	<	value += angMom[0]*angMom[0]/I(0, 0)
499	<	+ angMom[1]*angMom[1]/I(1, 1)
500	<	+ angMom[2]*angMom[2]/I(2, 2);
498	>	value += angMom[0]*angMom[0]/I(0, 0)
499	>	+ angMom[1]*angMom[1]/I(1, 1)
500	>	+ angMom[2]*angMom[2]/I(2, 2);
501		}
502	<	}
503	<	value = value * 0.5 / OOPSEConstant::energyConvert;
502	>	} //angular momenta exchange enabled
503	>	value *= 0.5;
504		break;
505		case rnemdPx :
506		value = mass * vel[0];
#	Line 202 | Line 511 | namespace oopse {
511		case rnemdPz :
512		value = mass * vel[2];
513		break;
205	–	case rnemdUnknown :
514		default :
515		break;
516		}
517
518	<	if (binNo == 0) {
518	>	if (inA == 0) {
519		if (!min_found) {
520		min_val = value;
521		min_sd = sd;
#	Line 218 | Line 526 | namespace oopse {
526		min_sd = sd;
527		}
528		}
529	<	} else {
529	>	} else {
530		if (!max_found) {
531		max_val = value;
532		max_sd = sd;
#	Line 232 | Line 540 | namespace oopse {
540		}
541		}
542		}
543	<
544	<	#ifdef IS_MPI
545	<	int nProc, worldRank;
546	<
239	<	nProc = MPI::COMM_WORLD.Get_size();
240	<	worldRank = MPI::COMM_WORLD.Get_rank();
241	<
543	>
544	>	#ifdef IS_MPI
545	>	int worldRank = MPI::COMM_WORLD.Get_rank();
546	>
547		bool my_min_found = min_found;
548		bool my_max_found = max_found;
549
550		// Even if we didn't find a minimum, did someone else?
551	<	MPI::COMM_WORLD.Allreduce(&my_min_found, &min_found,
247	<	1, MPI::BOOL, MPI::LAND);
248	<
551	>	MPI::COMM_WORLD.Allreduce(&my_min_found, &min_found, 1, MPI::BOOL, MPI::LOR);
552		// Even if we didn't find a maximum, did someone else?
553	<	MPI::COMM_WORLD.Allreduce(&my_max_found, &max_found,
554	<	1, MPI::BOOL, MPI::LAND);
555	<
556	<	struct {
557	<	RealType val;
558	<	int rank;
559	<	} max_vals, min_vals;
560	<
561	<	if (min_found) {
562	<	if (my_min_found)
553	>	MPI::COMM_WORLD.Allreduce(&my_max_found, &max_found, 1, MPI::BOOL, MPI::LOR);
554	>	#endif
555	>
556	>	if (max_found && min_found) {
557	>
558	>	#ifdef IS_MPI
559	>	struct {
560	>	RealType val;
561	>	int rank;
562	>	} max_vals, min_vals;
563	>
564	>	if (my_min_found) {
565		min_vals.val = min_val;
566	<	else
566	>	} else {
567		min_vals.val = HONKING_LARGE_VALUE;
568	<
568	>	}
569		min_vals.rank = worldRank;
570
571		// Who had the minimum?
572		MPI::COMM_WORLD.Allreduce(&min_vals, &min_vals,
573		1, MPI::REALTYPE_INT, MPI::MINLOC);
574		min_val = min_vals.val;
270	–	}
575
576	<	if (max_found) {
273	<	if (my_max_found)
576	>	if (my_max_found) {
577		max_vals.val = max_val;
578	<	else
578	>	} else {
579		max_vals.val = -HONKING_LARGE_VALUE;
580	<
580	>	}
581		max_vals.rank = worldRank;
582
583		// Who had the maximum?
584		MPI::COMM_WORLD.Allreduce(&max_vals, &max_vals,
585		1, MPI::REALTYPE_INT, MPI::MAXLOC);
586		max_val = max_vals.val;
284	–	}
587		#endif
588	<
589	<	if (max_found && min_found) {
590	<	if (min_val< max_val) {
289	<
588	>
589	>	if (min_val < max_val) {
590	>
591		#ifdef IS_MPI
592		if (max_vals.rank == worldRank && min_vals.rank == worldRank) {
593		// I have both maximum and minimum, so proceed like a single
594		// processor version:
595		#endif
596	<	// objects to be swapped: velocity & angular velocity
596	>
597		Vector3d min_vel = min_sd->getVel();
598		Vector3d max_vel = max_sd->getVel();
599		RealType temp_vel;
600
601	<	switch(rnemdType_) {
602	<	case rnemdKinetic :
601	>	switch(rnemdFluxType_) {
602	>	case rnemdKE :
603		min_sd->setVel(max_vel);
604		max_sd->setVel(min_vel);
605	<	if (min_sd->isDirectional() && max_sd->isDirectional()) {
605	>	if (min_sd->isDirectional() && max_sd->isDirectional()) {
606		Vector3d min_angMom = min_sd->getJ();
607		Vector3d max_angMom = max_sd->getJ();
608		min_sd->setJ(max_angMom);
609		max_sd->setJ(min_angMom);
610	<	}
610	>	}//angular momenta exchange enabled
611	>	//assumes same rigid body identity
612		break;
613		case rnemdPx :
614		temp_vel = min_vel.x();
#	Line 329 | Line 631 | namespace oopse {
631		min_sd->setVel(min_vel);
632		max_sd->setVel(max_vel);
633		break;
332	–	case rnemdUnknown :
634		default :
635		break;
636		}
637	+
638		#ifdef IS_MPI
639		// the rest of the cases only apply in parallel simulations:
640		} else if (max_vals.rank == worldRank) {
#	Line 348 | Line 650 | namespace oopse {
650		min_vel.getArrayPointer(), 3, MPI::REALTYPE,
651		min_vals.rank, 0, status);
652
653	<	switch(rnemdType_) {
654	<	case rnemdKinetic :
653	>	switch(rnemdFluxType_) {
654	>	case rnemdKE :
655		max_sd->setVel(min_vel);
656	<
656	>	//angular momenta exchange enabled
657		if (max_sd->isDirectional()) {
658		Vector3d min_angMom;
659		Vector3d max_angMom = max_sd->getJ();
660	<
660	>
661		// point-to-point swap of the angular momentum vector
662		MPI::COMM_WORLD.Sendrecv(max_angMom.getArrayPointer(), 3,
663		MPI::REALTYPE, min_vals.rank, 1,
664		min_angMom.getArrayPointer(), 3,
665		MPI::REALTYPE, min_vals.rank, 1,
666		status);
667	<
667	>
668		max_sd->setJ(min_angMom);
669	<	}
669	>	}
670		break;
671		case rnemdPx :
672		max_vel.x() = min_vel.x();
#	Line 378 | Line 680 | namespace oopse {
680		max_vel.z() = min_vel.z();
681		max_sd->setVel(max_vel);
682		break;
381	–	case rnemdUnknown :
683		default :
684		break;
685		}
#	Line 395 | Line 696 | namespace oopse {
696		max_vel.getArrayPointer(), 3, MPI::REALTYPE,
697		max_vals.rank, 0, status);
698
699	<	switch(rnemdType_) {
700	<	case rnemdKinetic :
699	>	switch(rnemdFluxType_) {
700	>	case rnemdKE :
701		min_sd->setVel(max_vel);
702	<
702	>	//angular momenta exchange enabled
703		if (min_sd->isDirectional()) {
704		Vector3d min_angMom = min_sd->getJ();
705		Vector3d max_angMom;
706	<
706	>
707		// point-to-point swap of the angular momentum vector
708		MPI::COMM_WORLD.Sendrecv(min_angMom.getArrayPointer(), 3,
709		MPI::REALTYPE, max_vals.rank, 1,
710		max_angMom.getArrayPointer(), 3,
711		MPI::REALTYPE, max_vals.rank, 1,
712		status);
713	<
713	>
714		min_sd->setJ(max_angMom);
715		}
716		break;
#	Line 425 | Line 726 | namespace oopse {
726		min_vel.z() = max_vel.z();
727		min_sd->setVel(min_vel);
728		break;
428	–	case rnemdUnknown :
729		default :
730		break;
731		}
732		}
733		#endif
734	<	exchangeSum_ += max_val - min_val;
735	<	} else {
736	<	std::cerr << "exchange NOT performed.\nmin_val > max_val.\n";
734	>
735	>	switch(rnemdFluxType_) {
736	>	case rnemdKE:
737	>	kineticExchange_ += max_val - min_val;
738	>	break;
739	>	case rnemdPx:
740	>	momentumExchange_.x() += max_val - min_val;
741	>	break;
742	>	case rnemdPy:
743	>	momentumExchange_.y() += max_val - min_val;
744	>	break;
745	>	case rnemdPz:
746	>	momentumExchange_.z() += max_val - min_val;
747	>	break;
748	>	default:
749	>	break;
750	>	}
751	>	} else {
752	>	sprintf(painCave.errMsg,
753	>	"RNEMD::doSwap exchange NOT performed because min_val > max_val\n");
754	>	painCave.isFatal = 0;
755	>	painCave.severity = OPENMD_INFO;
756	>	simError();
757	>	failTrialCount_++;
758		}
759		} else {
760	<	std::cerr << "exchange NOT performed.\none of the two slabs empty.\n";
761	<	}
762	<
760	>	sprintf(painCave.errMsg,
761	>	"RNEMD::doSwap exchange NOT performed because selected object\n"
762	>	"\twas not present in at least one of the two slabs.\n");
763	>	painCave.isFatal = 0;
764	>	painCave.severity = OPENMD_INFO;
765	>	simError();
766	>	failTrialCount_++;
767	>	}
768		}
769
770	<	void RNEMD::getStatus() {
771	<
770	>	void RNEMD::doNIVS() {
771	>	if (!doRNEMD_) return;
772		Snapshot* currentSnap_ = info_->getSnapshotManager()->getCurrentSnapshot();
773		Mat3x3d hmat = currentSnap_->getHmat();
448	–	Stats& stat = currentSnap_->statData;
449	–	RealType time = currentSnap_->getTime();
774
451	–	stat[Stats::RNEMD_SWAP_TOTAL] = exchangeSum_;
452	–
775		seleMan_.setSelectionSet(evaluator_.evaluate());
776
777		int selei;
778		StuntDouble* sd;
457	–	int idx;
779
780	<	std::vector<RealType> valueHist(nBins_, 0.0); // keeps track of what's
460	<	// being averaged
461	<	std::vector<int> valueCount(nBins_, 0);       // keeps track of the
462	<	// number of degrees of
463	<	// freedom being averaged
780	>	vector<StuntDouble*> hotBin, coldBin;
781
782	+	RealType Phx = 0.0;
783	+	RealType Phy = 0.0;
784	+	RealType Phz = 0.0;
785	+	RealType Khx = 0.0;
786	+	RealType Khy = 0.0;
787	+	RealType Khz = 0.0;
788	+	RealType Khw = 0.0;
789	+	RealType Pcx = 0.0;
790	+	RealType Pcy = 0.0;
791	+	RealType Pcz = 0.0;
792	+	RealType Kcx = 0.0;
793	+	RealType Kcy = 0.0;
794	+	RealType Kcz = 0.0;
795	+	RealType Kcw = 0.0;
796	+
797		for (sd = seleMan_.beginSelected(selei); sd != NULL;
798		sd = seleMan_.nextSelected(selei)) {
799	<
468	<	idx = sd->getLocalIndex();
469	<
799	>
800		Vector3d pos = sd->getPos();
801
802		// wrap the stuntdouble's position back into the box:
803	<
803	>
804		if (usePeriodicBoundaryConditions_)
805		currentSnap_->wrapVector(pos);
806	<
806	>
807		// which bin is this stuntdouble in?
808	<	// wrapped positions are in the range [-0.5*hmat(2,2), +0.5*hmat(2,2)]
809	<
480	<	int binNo = int(nBins_ * (pos.z() / hmat(2,2) + 0.5)) % nBins_;
481	<
482	<	RealType mass = sd->getMass();
483	<	Vector3d vel = sd->getVel();
484	<	RealType value;
808	>	bool inA = inSlabA(pos);
809	>	bool inB = inSlabB(pos);
810
811	<	switch(rnemdType_) {
812	<	case rnemdKinetic :
813	<
814	<	value = mass * (vel[0]*vel[0] + vel[1]*vel[1] +
815	<	vel[2]*vel[2]);
816	<
817	<	valueCount[binNo] += 3;
818	<	if (sd->isDirectional()) {
819	<	Vector3d angMom = sd->getJ();
820	<	Mat3x3d I = sd->getI();
821	<
822	<	if (sd->isLinear()) {
823	<	int i = sd->linearAxis();
824	<	int j = (i + 1) % 3;
825	<	int k = (i + 2) % 3;
826	<	value += angMom[j] * angMom[j] / I(j, j) +
827	<	angMom[k] * angMom[k] / I(k, k);
811	>	if (inA || inB) {
812	>
813	>	RealType mass = sd->getMass();
814	>	Vector3d vel = sd->getVel();
815	>
816	>	if (inA) {
817	>	hotBin.push_back(sd);
818	>	Phx += mass * vel.x();
819	>	Phy += mass * vel.y();
820	>	Phz += mass * vel.z();
821	>	Khx += mass * vel.x() * vel.x();
822	>	Khy += mass * vel.y() * vel.y();
823	>	Khz += mass * vel.z() * vel.z();
824	>	if (sd->isDirectional()) {
825	>	Vector3d angMom = sd->getJ();
826	>	Mat3x3d I = sd->getI();
827	>	if (sd->isLinear()) {
828	>	int i = sd->linearAxis();
829	>	int j = (i + 1) % 3;
830	>	int k = (i + 2) % 3;
831	>	Khw += angMom[j] * angMom[j] / I(j, j) +
832	>	angMom[k] * angMom[k] / I(k, k);
833	>	} else {
834	>	Khw += angMom[0]*angMom[0]/I(0, 0)
835	>	+ angMom[1]*angMom[1]/I(1, 1)
836	>	+ angMom[2]*angMom[2]/I(2, 2);
837	>	}
838	>	}
839	>	} else {
840	>	coldBin.push_back(sd);
841	>	Pcx += mass * vel.x();
842	>	Pcy += mass * vel.y();
843	>	Pcz += mass * vel.z();
844	>	Kcx += mass * vel.x() * vel.x();
845	>	Kcy += mass * vel.y() * vel.y();
846	>	Kcz += mass * vel.z() * vel.z();
847	>	if (sd->isDirectional()) {
848	>	Vector3d angMom = sd->getJ();
849	>	Mat3x3d I = sd->getI();
850	>	if (sd->isLinear()) {
851	>	int i = sd->linearAxis();
852	>	int j = (i + 1) % 3;
853	>	int k = (i + 2) % 3;
854	>	Kcw += angMom[j] * angMom[j] / I(j, j) +
855	>	angMom[k] * angMom[k] / I(k, k);
856	>	} else {
857	>	Kcw += angMom[0]*angMom[0]/I(0, 0)
858	>	+ angMom[1]*angMom[1]/I(1, 1)
859	>	+ angMom[2]*angMom[2]/I(2, 2);
860	>	}
861	>	}
862	>	}
863	>	}
864	>	}
865	>
866	>	Khx *= 0.5;
867	>	Khy *= 0.5;
868	>	Khz *= 0.5;
869	>	Khw *= 0.5;
870	>	Kcx *= 0.5;
871	>	Kcy *= 0.5;
872	>	Kcz *= 0.5;
873	>	Kcw *= 0.5;
874
875	<	valueCount[binNo] +=2;
875	>	#ifdef IS_MPI
876	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Phx, 1, MPI::REALTYPE, MPI::SUM);
877	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Phy, 1, MPI::REALTYPE, MPI::SUM);
878	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Phz, 1, MPI::REALTYPE, MPI::SUM);
879	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Pcx, 1, MPI::REALTYPE, MPI::SUM);
880	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Pcy, 1, MPI::REALTYPE, MPI::SUM);
881	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Pcz, 1, MPI::REALTYPE, MPI::SUM);
882
883	<	} else {
884	<	value += angMom[0]*angMom[0]/I(0, 0)
885	<	+ angMom[1]*angMom[1]/I(1, 1)
886	<	+ angMom[2]*angMom[2]/I(2, 2);
887	<	valueCount[binNo] +=3;
883	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Khx, 1, MPI::REALTYPE, MPI::SUM);
884	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Khy, 1, MPI::REALTYPE, MPI::SUM);
885	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Khz, 1, MPI::REALTYPE, MPI::SUM);
886	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Khw, 1, MPI::REALTYPE, MPI::SUM);
887	>
888	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Kcx, 1, MPI::REALTYPE, MPI::SUM);
889	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Kcy, 1, MPI::REALTYPE, MPI::SUM);
890	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Kcz, 1, MPI::REALTYPE, MPI::SUM);
891	>	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Kcw, 1, MPI::REALTYPE, MPI::SUM);
892	>	#endif
893	>
894	>	//solve coldBin coeff's first
895	>	RealType px = Pcx / Phx;
896	>	RealType py = Pcy / Phy;
897	>	RealType pz = Pcz / Phz;
898	>	RealType c, x, y, z;
899	>	bool successfulScale = false;
900	>	if ((rnemdFluxType_ == rnemdFullKE) ||
901	>	(rnemdFluxType_ == rnemdRotKE)) {
902	>	//may need sanity check Khw & Kcw > 0
903	>
904	>	if (rnemdFluxType_ == rnemdFullKE) {
905	>	c = 1.0 - kineticTarget_ / (Kcx + Kcy + Kcz + Kcw);
906	>	} else {
907	>	c = 1.0 - kineticTarget_ / Kcw;
908	>	}
909	>
910	>	if ((c > 0.81) && (c < 1.21)) {//restrict scaling coefficients
911	>	c = sqrt(c);
912	>
913	>	RealType w = 0.0;
914	>	if (rnemdFluxType_ ==  rnemdFullKE) {
915	>	x = 1.0 + px * (1.0 - c);
916	>	y = 1.0 + py * (1.0 - c);
917	>	z = 1.0 + pz * (1.0 - c);
918	>	/* more complicated way
919	>	w = 1.0 + (Kcw - Kcw * c * c - (c * c * (Kcx + Kcy + Kcz
920	>	+ Khx * px * px + Khy * py * py + Khz * pz * pz)
921	>	- 2.0 * c * (Khx * px * (1.0 + px) + Khy * py * (1.0 + py)
922	>	+ Khz * pz * (1.0 + pz)) + Khx * px * (2.0 + px)
923	>	+ Khy * py * (2.0 + py) + Khz * pz * (2.0 + pz)
924	>	- Kcx - Kcy - Kcz)) / Khw; the following is simpler
925	>	*/
926	>	if ((fabs(x - 1.0) < 0.1) && (fabs(y - 1.0) < 0.1) &&
927	>	(fabs(z - 1.0) < 0.1)) {
928	>	w = 1.0 + (kineticTarget_
929	>	+ Khx * (1.0 - x * x) + Khy * (1.0 - y * y)
930	>	+ Khz * (1.0 - z * z)) / Khw;
931	>	}//no need to calculate w if x, y or z is out of range
932	>	} else {
933	>	w = 1.0 + kineticTarget_ / Khw;
934	>	}
935	>	if ((w > 0.81) && (w < 1.21)) {//restrict scaling coefficients
936	>	//if w is in the right range, so should be x, y, z.
937	>	vector<StuntDouble*>::iterator sdi;
938	>	Vector3d vel;
939	>	for (sdi = coldBin.begin(); sdi != coldBin.end(); sdi++) {
940	>	if (rnemdFluxType_ == rnemdFullKE) {
941	>	vel = (*sdi)->getVel() * c;
942	>	(*sdi)->setVel(vel);
943	>	}
944	>	if ((*sdi)->isDirectional()) {
945	>	Vector3d angMom = (*sdi)->getJ() * c;
946	>	(*sdi)->setJ(angMom);
947	>	}
948		}
949	+	w = sqrt(w);
950	+	for (sdi = hotBin.begin(); sdi != hotBin.end(); sdi++) {
951	+	if (rnemdFluxType_ == rnemdFullKE) {
952	+	vel = (*sdi)->getVel();
953	+	vel.x() *= x;
954	+	vel.y() *= y;
955	+	vel.z() *= z;
956	+	(*sdi)->setVel(vel);
957	+	}
958	+	if ((*sdi)->isDirectional()) {
959	+	Vector3d angMom = (*sdi)->getJ() * w;
960	+	(*sdi)->setJ(angMom);
961	+	}
962	+	}
963	+	successfulScale = true;
964	+	kineticExchange_ += kineticTarget_;
965		}
966	<	value = value / OOPSEConstant::energyConvert / OOPSEConstant::kb;
967	<
968	<	break;
969	<	case rnemdPx :
970	<	value = mass * vel[0];
971	<	valueCount[binNo]++;
966	>	}
967	>	} else {
968	>	RealType a000, a110, c0, a001, a111, b01, b11, c1;
969	>	switch(rnemdFluxType_) {
970	>	case rnemdKE :
971	>	/* used hotBin coeff's & only scale x & y dimensions
972	>	RealType px = Phx / Pcx;
973	>	RealType py = Phy / Pcy;
974	>	a110 = Khy;
975	>	c0 = - Khx - Khy - kineticTarget_;
976	>	a000 = Khx;
977	>	a111 = Kcy * py * py;
978	>	b11 = -2.0 * Kcy * py * (1.0 + py);
979	>	c1 = Kcy * py * (2.0 + py) + Kcx * px * ( 2.0 + px) + kineticTarget_;
980	>	b01 = -2.0 * Kcx * px * (1.0 + px);
981	>	a001 = Kcx * px * px;
982	>	*/
983	>	//scale all three dimensions, let c_x = c_y
984	>	a000 = Kcx + Kcy;
985	>	a110 = Kcz;
986	>	c0 = kineticTarget_ - Kcx - Kcy - Kcz;
987	>	a001 = Khx * px * px + Khy * py * py;
988	>	a111 = Khz * pz * pz;
989	>	b01 = -2.0 * (Khx * px * (1.0 + px) + Khy * py * (1.0 + py));
990	>	b11 = -2.0 * Khz * pz * (1.0 + pz);
991	>	c1 = Khx * px * (2.0 + px) + Khy * py * (2.0 + py)
992	>	+ Khz * pz * (2.0 + pz) - kineticTarget_;
993		break;
994	+	case rnemdPx :
995	+	c = 1 - momentumTarget_.x() / Pcx;
996	+	a000 = Kcy;
997	+	a110 = Kcz;
998	+	c0 = Kcx * c * c - Kcx - Kcy - Kcz;
999	+	a001 = py * py * Khy;
1000	+	a111 = pz * pz * Khz;
1001	+	b01 = -2.0 * Khy * py * (1.0 + py);
1002	+	b11 = -2.0 * Khz * pz * (1.0 + pz);
1003	+	c1 = Khy * py * (2.0 + py) + Khz * pz * (2.0 + pz)
1004	+	+ Khx * (fastpow(c * px - px - 1.0, 2) - 1.0);
1005	+	break;
1006		case rnemdPy :
1007	<	value = mass * vel[1];
1008	<	valueCount[binNo]++;
1007	>	c = 1 - momentumTarget_.y() / Pcy;
1008	>	a000 = Kcx;
1009	>	a110 = Kcz;
1010	>	c0 = Kcy * c * c - Kcx - Kcy - Kcz;
1011	>	a001 = px * px * Khx;
1012	>	a111 = pz * pz * Khz;
1013	>	b01 = -2.0 * Khx * px * (1.0 + px);
1014	>	b11 = -2.0 * Khz * pz * (1.0 + pz);
1015	>	c1 = Khx * px * (2.0 + px) + Khz * pz * (2.0 + pz)
1016	>	+ Khy * (fastpow(c * py - py - 1.0, 2) - 1.0);
1017		break;
1018	<	case rnemdPz :
1019	<	value = mass * vel[2];
1020	<	valueCount[binNo]++;
1018	>	case rnemdPz ://we don't really do this, do we?
1019	>	c = 1 - momentumTarget_.z() / Pcz;
1020	>	a000 = Kcx;
1021	>	a110 = Kcy;
1022	>	c0 = Kcz * c * c - Kcx - Kcy - Kcz;
1023	>	a001 = px * px * Khx;
1024	>	a111 = py * py * Khy;
1025	>	b01 = -2.0 * Khx * px * (1.0 + px);
1026	>	b11 = -2.0 * Khy * py * (1.0 + py);
1027	>	c1 = Khx * px * (2.0 + px) + Khy * py * (2.0 + py)
1028	>	+ Khz * (fastpow(c * pz - pz - 1.0, 2) - 1.0);
1029		break;
528	–	case rnemdUnknown :
1030		default :
1031		break;
1032		}
1033	<	valueHist[binNo] += value;
1033	>
1034	>	RealType v1 = a000 * a111 - a001 * a110;
1035	>	RealType v2 = a000 * b01;
1036	>	RealType v3 = a000 * b11;
1037	>	RealType v4 = a000 * c1 - a001 * c0;
1038	>	RealType v8 = a110 * b01;
1039	>	RealType v10 = - b01 * c0;
1040	>
1041	>	RealType u0 = v2 * v10 - v4 * v4;
1042	>	RealType u1 = -2.0 * v3 * v4;
1043	>	RealType u2 = -v2 * v8 - v3 * v3 - 2.0 * v1 * v4;
1044	>	RealType u3 = -2.0 * v1 * v3;
1045	>	RealType u4 = - v1 * v1;
1046	>	//rescale coefficients
1047	>	RealType maxAbs = fabs(u0);
1048	>	if (maxAbs < fabs(u1)) maxAbs = fabs(u1);
1049	>	if (maxAbs < fabs(u2)) maxAbs = fabs(u2);
1050	>	if (maxAbs < fabs(u3)) maxAbs = fabs(u3);
1051	>	if (maxAbs < fabs(u4)) maxAbs = fabs(u4);
1052	>	u0 /= maxAbs;
1053	>	u1 /= maxAbs;
1054	>	u2 /= maxAbs;
1055	>	u3 /= maxAbs;
1056	>	u4 /= maxAbs;
1057	>	//max_element(start, end) is also available.
1058	>	Polynomial<RealType> poly; //same as DoublePolynomial poly;
1059	>	poly.setCoefficient(4, u4);
1060	>	poly.setCoefficient(3, u3);
1061	>	poly.setCoefficient(2, u2);
1062	>	poly.setCoefficient(1, u1);
1063	>	poly.setCoefficient(0, u0);
1064	>	vector<RealType> realRoots = poly.FindRealRoots();
1065	>
1066	>	vector<RealType>::iterator ri;
1067	>	RealType r1, r2, alpha0;
1068	>	vector<pair<RealType,RealType> > rps;
1069	>	for (ri = realRoots.begin(); ri !=realRoots.end(); ri++) {
1070	>	r2 = *ri;
1071	>	//check if FindRealRoots() give the right answer
1072	>	if ( fabs(u0 + r2 * (u1 + r2 * (u2 + r2 * (u3 + r2 * u4)))) > 1e-6 ) {
1073	>	sprintf(painCave.errMsg,
1074	>	"RNEMD Warning: polynomial solve seems to have an error!");
1075	>	painCave.isFatal = 0;
1076	>	simError();
1077	>	failRootCount_++;
1078	>	}
1079	>	//might not be useful w/o rescaling coefficients
1080	>	alpha0 = -c0 - a110 * r2 * r2;
1081	>	if (alpha0 >= 0.0) {
1082	>	r1 = sqrt(alpha0 / a000);
1083	>	if (fabs(c1 + r1 * (b01 + r1 * a001) + r2 * (b11 + r2 * a111))
1084	>	< 1e-6)
1085	>	{ rps.push_back(make_pair(r1, r2)); }
1086	>	if (r1 > 1e-6) { //r1 non-negative
1087	>	r1 = -r1;
1088	>	if (fabs(c1 + r1 * (b01 + r1 * a001) + r2 * (b11 + r2 * a111))
1089	>	< 1e-6)
1090	>	{ rps.push_back(make_pair(r1, r2)); }
1091	>	}
1092	>	}
1093	>	}
1094	>	// Consider combining together the solving pair part w/ the searching
1095	>	// best solution part so that we don't need the pairs vector
1096	>	if (!rps.empty()) {
1097	>	RealType smallestDiff = HONKING_LARGE_VALUE;
1098	>	RealType diff;
1099	>	pair<RealType,RealType> bestPair = make_pair(1.0, 1.0);
1100	>	vector<pair<RealType,RealType> >::iterator rpi;
1101	>	for (rpi = rps.begin(); rpi != rps.end(); rpi++) {
1102	>	r1 = (*rpi).first;
1103	>	r2 = (*rpi).second;
1104	>	switch(rnemdFluxType_) {
1105	>	case rnemdKE :
1106	>	diff = fastpow(1.0 - r1, 2) + fastpow(1.0 - r2, 2)
1107	>	+ fastpow(r1 * r1 / r2 / r2 - Kcz/Kcx, 2)
1108	>	+ fastpow(r1 * r1 / r2 / r2 - Kcz/Kcy, 2);
1109	>	break;
1110	>	case rnemdPx :
1111	>	diff = fastpow(1.0 - r1, 2) + fastpow(1.0 - r2, 2)
1112	>	+ fastpow(r1 * r1 / r2 / r2 - Kcz/Kcy, 2);
1113	>	break;
1114	>	case rnemdPy :
1115	>	diff = fastpow(1.0 - r1, 2) + fastpow(1.0 - r2, 2)
1116	>	+ fastpow(r1 * r1 / r2 / r2 - Kcz/Kcx, 2);
1117	>	break;
1118	>	case rnemdPz :
1119	>	diff = fastpow(1.0 - r1, 2) + fastpow(1.0 - r2, 2)
1120	>	+ fastpow(r1 * r1 / r2 / r2 - Kcy/Kcx, 2);
1121	>	default :
1122	>	break;
1123	>	}
1124	>	if (diff < smallestDiff) {
1125	>	smallestDiff = diff;
1126	>	bestPair = *rpi;
1127	>	}
1128	>	}
1129	>	#ifdef IS_MPI
1130	>	if (worldRank == 0) {
1131	>	#endif
1132	>	// sprintf(painCave.errMsg,
1133	>	//         "RNEMD: roots r1= %lf\tr2 = %lf\n",
1134	>	//         bestPair.first, bestPair.second);
1135	>	// painCave.isFatal = 0;
1136	>	// painCave.severity = OPENMD_INFO;
1137	>	// simError();
1138	>	#ifdef IS_MPI
1139	>	}
1140	>	#endif
1141	>
1142	>	switch(rnemdFluxType_) {
1143	>	case rnemdKE :
1144	>	x = bestPair.first;
1145	>	y = bestPair.first;
1146	>	z = bestPair.second;
1147	>	break;
1148	>	case rnemdPx :
1149	>	x = c;
1150	>	y = bestPair.first;
1151	>	z = bestPair.second;
1152	>	break;
1153	>	case rnemdPy :
1154	>	x = bestPair.first;
1155	>	y = c;
1156	>	z = bestPair.second;
1157	>	break;
1158	>	case rnemdPz :
1159	>	x = bestPair.first;
1160	>	y = bestPair.second;
1161	>	z = c;
1162	>	break;
1163	>	default :
1164	>	break;
1165	>	}
1166	>	vector<StuntDouble*>::iterator sdi;
1167	>	Vector3d vel;
1168	>	for (sdi = coldBin.begin(); sdi != coldBin.end(); sdi++) {
1169	>	vel = (*sdi)->getVel();
1170	>	vel.x() *= x;
1171	>	vel.y() *= y;
1172	>	vel.z() *= z;
1173	>	(*sdi)->setVel(vel);
1174	>	}
1175	>	//convert to hotBin coefficient
1176	>	x = 1.0 + px * (1.0 - x);
1177	>	y = 1.0 + py * (1.0 - y);
1178	>	z = 1.0 + pz * (1.0 - z);
1179	>	for (sdi = hotBin.begin(); sdi != hotBin.end(); sdi++) {
1180	>	vel = (*sdi)->getVel();
1181	>	vel.x() *= x;
1182	>	vel.y() *= y;
1183	>	vel.z() *= z;
1184	>	(*sdi)->setVel(vel);
1185	>	}
1186	>	successfulScale = true;
1187	>	switch(rnemdFluxType_) {
1188	>	case rnemdKE :
1189	>	kineticExchange_ += kineticTarget_;
1190	>	break;
1191	>	case rnemdPx :
1192	>	case rnemdPy :
1193	>	case rnemdPz :
1194	>	momentumExchange_ += momentumTarget_;
1195	>	break;
1196	>	default :
1197	>	break;
1198	>	}
1199	>	}
1200		}
1201	+	if (successfulScale != true) {
1202	+	sprintf(painCave.errMsg,
1203	+	"RNEMD::doNIVS exchange NOT performed - roots that solve\n"
1204	+	"\tthe constraint equations may not exist or there may be\n"
1205	+	"\tno selected objects in one or both slabs.\n");
1206	+	painCave.isFatal = 0;
1207	+	painCave.severity = OPENMD_INFO;
1208	+	simError();
1209	+	failTrialCount_++;
1210	+	}
1211	+	}
1212
1213	+	void RNEMD::doVSS() {
1214	+	if (!doRNEMD_) return;
1215	+	Snapshot* currentSnap_ = info_->getSnapshotManager()->getCurrentSnapshot();
1216	+	RealType time = currentSnap_->getTime();
1217	+	Mat3x3d hmat = currentSnap_->getHmat();
1218	+
1219	+	seleMan_.setSelectionSet(evaluator_.evaluate());
1220	+
1221	+	int selei;
1222	+	StuntDouble* sd;
1223	+
1224	+	vector<StuntDouble*> hotBin, coldBin;
1225	+
1226	+	Vector3d Ph(V3Zero);
1227	+	RealType Mh = 0.0;
1228	+	RealType Kh = 0.0;
1229	+	Vector3d Pc(V3Zero);
1230	+	RealType Mc = 0.0;
1231	+	RealType Kc = 0.0;
1232	+
1233	+
1234	+	for (sd = seleMan_.beginSelected(selei); sd != NULL;
1235	+	sd = seleMan_.nextSelected(selei)) {
1236	+
1237	+	Vector3d pos = sd->getPos();
1238	+
1239	+	// wrap the stuntdouble's position back into the box:
1240	+
1241	+	if (usePeriodicBoundaryConditions_)
1242	+	currentSnap_->wrapVector(pos);
1243	+
1244	+	// which bin is this stuntdouble in?
1245	+	bool inA = inSlabA(pos);
1246	+	bool inB = inSlabB(pos);
1247	+
1248	+	if (inA || inB) {
1249	+
1250	+	RealType mass = sd->getMass();
1251	+	Vector3d vel = sd->getVel();
1252	+
1253	+	if (inA) {
1254	+	hotBin.push_back(sd);
1255	+	Ph += mass * vel;
1256	+	Mh += mass;
1257	+	Kh += mass * vel.lengthSquare();
1258	+	if (rnemdFluxType_ == rnemdFullKE) {
1259	+	if (sd->isDirectional()) {
1260	+	Vector3d angMom = sd->getJ();
1261	+	Mat3x3d I = sd->getI();
1262	+	if (sd->isLinear()) {
1263	+	int i = sd->linearAxis();
1264	+	int j = (i + 1) % 3;
1265	+	int k = (i + 2) % 3;
1266	+	Kh += angMom[j] * angMom[j] / I(j, j) +
1267	+	angMom[k] * angMom[k] / I(k, k);
1268	+	} else {
1269	+	Kh += angMom[0] * angMom[0] / I(0, 0) +
1270	+	angMom[1] * angMom[1] / I(1, 1) +
1271	+	angMom[2] * angMom[2] / I(2, 2);
1272	+	}
1273	+	}
1274	+	}
1275	+	} else { //midBin_
1276	+	coldBin.push_back(sd);
1277	+	Pc += mass * vel;
1278	+	Mc += mass;
1279	+	Kc += mass * vel.lengthSquare();
1280	+	if (rnemdFluxType_ == rnemdFullKE) {
1281	+	if (sd->isDirectional()) {
1282	+	Vector3d angMom = sd->getJ();
1283	+	Mat3x3d I = sd->getI();
1284	+	if (sd->isLinear()) {
1285	+	int i = sd->linearAxis();
1286	+	int j = (i + 1) % 3;
1287	+	int k = (i + 2) % 3;
1288	+	Kc += angMom[j] * angMom[j] / I(j, j) +
1289	+	angMom[k] * angMom[k] / I(k, k);
1290	+	} else {
1291	+	Kc += angMom[0] * angMom[0] / I(0, 0) +
1292	+	angMom[1] * angMom[1] / I(1, 1) +
1293	+	angMom[2] * angMom[2] / I(2, 2);
1294	+	}
1295	+	}
1296	+	}
1297	+	}
1298	+	}
1299	+	}
1300	+
1301	+	Kh *= 0.5;
1302	+	Kc *= 0.5;
1303	+
1304		#ifdef IS_MPI
1305	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Ph[0], 3, MPI::REALTYPE, MPI::SUM);
1306	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Pc[0], 3, MPI::REALTYPE, MPI::SUM);
1307	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Mh, 1, MPI::REALTYPE, MPI::SUM);
1308	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Kh, 1, MPI::REALTYPE, MPI::SUM);
1309	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Mc, 1, MPI::REALTYPE, MPI::SUM);
1310	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &Kc, 1, MPI::REALTYPE, MPI::SUM);
1311	+	#endif
1312
1313	<	// all processors have the same number of bins, and STL vectors pack their
1314	<	// arrays, so in theory, this should be safe:
1313	>	bool successfulExchange = false;
1314	>	if ((Mh > 0.0) && (Mc > 0.0)) {//both slabs are not empty
1315	>	Vector3d vc = Pc / Mc;
1316	>	Vector3d ac = -momentumTarget_ / Mc + vc;
1317	>	Vector3d acrec = -momentumTarget_ / Mc;
1318	>	RealType cNumerator = Kc - kineticTarget_ - 0.5 * Mc * ac.lengthSquare();
1319	>	if (cNumerator > 0.0) {
1320	>	RealType cDenominator = Kc - 0.5 * Mc * vc.lengthSquare();
1321	>	if (cDenominator > 0.0) {
1322	>	RealType c = sqrt(cNumerator / cDenominator);
1323	>	if ((c > 0.9) && (c < 1.1)) {//restrict scaling coefficients
1324	>	Vector3d vh = Ph / Mh;
1325	>	Vector3d ah = momentumTarget_ / Mh + vh;
1326	>	Vector3d ahrec = momentumTarget_ / Mh;
1327	>	RealType hNumerator = Kh + kineticTarget_
1328	>	- 0.5 * Mh * ah.lengthSquare();
1329	>	if (hNumerator > 0.0) {
1330	>	RealType hDenominator = Kh - 0.5 * Mh * vh.lengthSquare();
1331	>	if (hDenominator > 0.0) {
1332	>	RealType h = sqrt(hNumerator / hDenominator);
1333	>	if ((h > 0.9) && (h < 1.1)) {
1334
1335	<	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &valueHist[0],
1336	<	nBins_, MPI::REALTYPE, MPI::SUM);
1337	<	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &valueCount[0],
1338	<	nBins_, MPI::INT, MPI::SUM);
1335	>	vector<StuntDouble*>::iterator sdi;
1336	>	Vector3d vel;
1337	>	for (sdi = coldBin.begin(); sdi != coldBin.end(); sdi++) {
1338	>	//vel = (*sdi)->getVel();
1339	>	vel = ((*sdi)->getVel() - vc) * c + ac;
1340	>	(*sdi)->setVel(vel);
1341	>	if (rnemdFluxType_ == rnemdFullKE) {
1342	>	if ((*sdi)->isDirectional()) {
1343	>	Vector3d angMom = (*sdi)->getJ() * c;
1344	>	(*sdi)->setJ(angMom);
1345	>	}
1346	>	}
1347	>	}
1348	>	for (sdi = hotBin.begin(); sdi != hotBin.end(); sdi++) {
1349	>	//vel = (*sdi)->getVel();
1350	>	vel = ((*sdi)->getVel() - vh) * h + ah;
1351	>	(*sdi)->setVel(vel);
1352	>	if (rnemdFluxType_ == rnemdFullKE) {
1353	>	if ((*sdi)->isDirectional()) {
1354	>	Vector3d angMom = (*sdi)->getJ() * h;
1355	>	(*sdi)->setJ(angMom);
1356	>	}
1357	>	}
1358	>	}
1359	>	successfulExchange = true;
1360	>	kineticExchange_ += kineticTarget_;
1361	>	momentumExchange_ += momentumTarget_;
1362	>	}
1363	>	}
1364	>	}
1365	>	}
1366	>	}
1367	>	}
1368	>	}
1369	>	if (successfulExchange != true) {
1370	>	sprintf(painCave.errMsg,
1371	>	"RNEMD::doVSS exchange NOT performed - roots that solve\n"
1372	>	"\tthe constraint equations may not exist or there may be\n"
1373	>	"\tno selected objects in one or both slabs.\n");
1374	>	painCave.isFatal = 0;
1375	>	painCave.severity = OPENMD_INFO;
1376	>	simError();
1377	>	failTrialCount_++;
1378	>	}
1379	>	}
1380	>
1381	>	void RNEMD::doRNEMD() {
1382	>	if (!doRNEMD_) return;
1383	>	trialCount_++;
1384	>	switch(rnemdMethod_) {
1385	>	case rnemdSwap:
1386	>	doSwap();
1387	>	break;
1388	>	case rnemdNIVS:
1389	>	doNIVS();
1390	>	break;
1391	>	case rnemdVSS:
1392	>	doVSS();
1393	>	break;
1394	>	case rnemdUnkownMethod:
1395	>	default :
1396	>	break;
1397	>	}
1398	>	}
1399
1400	+	void RNEMD::collectData() {
1401	+	if (!doRNEMD_) return;
1402	+	Snapshot* currentSnap_ = info_->getSnapshotManager()->getCurrentSnapshot();
1403	+	Mat3x3d hmat = currentSnap_->getHmat();
1404	+
1405	+	areaAccumulator_->add(currentSnap_->getXYarea());
1406	+
1407	+	seleMan_.setSelectionSet(evaluator_.evaluate());
1408	+
1409	+	int selei(0);
1410	+	StuntDouble* sd;
1411	+
1412	+	vector<RealType> binMass(nBins_, 0.0);
1413	+	vector<RealType> binPx(nBins_, 0.0);
1414	+	vector<RealType> binPy(nBins_, 0.0);
1415	+	vector<RealType> binPz(nBins_, 0.0);
1416	+	vector<RealType> binKE(nBins_, 0.0);
1417	+	vector<int> binDOF(nBins_, 0);
1418	+	vector<int> binCount(nBins_, 0);
1419	+
1420	+	// alternative approach, track all molecules instead of only those
1421	+	// selected for scaling/swapping:
1422	+	/*
1423	+	SimInfo::MoleculeIterator miter;
1424	+	vector<StuntDouble*>::iterator iiter;
1425	+	Molecule* mol;
1426	+	StuntDouble* sd;
1427	+	for (mol = info_->beginMolecule(miter); mol != NULL;
1428	+	mol = info_->nextMolecule(miter))
1429	+	sd is essentially sd
1430	+	for (sd = mol->beginIntegrableObject(iiter);
1431	+	sd != NULL;
1432	+	sd = mol->nextIntegrableObject(iiter))
1433	+	*/
1434	+
1435	+	for (sd = seleMan_.beginSelected(selei); sd != NULL;
1436	+	sd = seleMan_.nextSelected(selei)) {
1437	+
1438	+	Vector3d pos = sd->getPos();
1439	+
1440	+	// wrap the stuntdouble's position back into the box:
1441	+
1442	+	if (usePeriodicBoundaryConditions_)
1443	+	currentSnap_->wrapVector(pos);
1444	+
1445	+
1446	+	// which bin is this stuntdouble in?
1447	+	// wrapped positions are in the range [-0.5*hmat(2,2), +0.5*hmat(2,2)]
1448	+	// Shift molecules by half a box to have bins start at 0
1449	+	// The modulo operator is used to wrap the case when we are
1450	+	// beyond the end of the bins back to the beginning.
1451	+	int binNo = int(nBins_ * (pos.z() / hmat(2,2) + 0.5)) % nBins_;
1452	+
1453	+	RealType mass = sd->getMass();
1454	+	Vector3d vel = sd->getVel();
1455	+
1456	+	binCount[binNo]++;
1457	+	binMass[binNo] += mass;
1458	+	binPx[binNo] += mass*vel.x();
1459	+	binPy[binNo] += mass*vel.y();
1460	+	binPz[binNo] += mass*vel.z();
1461	+	binKE[binNo] += 0.5 * (mass * vel.lengthSquare());
1462	+	binDOF[binNo] += 3;
1463	+
1464	+	if (sd->isDirectional()) {
1465	+	Vector3d angMom = sd->getJ();
1466	+	Mat3x3d I = sd->getI();
1467	+	if (sd->isLinear()) {
1468	+	int i = sd->linearAxis();
1469	+	int j = (i + 1) % 3;
1470	+	int k = (i + 2) % 3;
1471	+	binKE[binNo] += 0.5 * (angMom[j] * angMom[j] / I(j, j) +
1472	+	angMom[k] * angMom[k] / I(k, k));
1473	+	binDOF[binNo] += 2;
1474	+	} else {
1475	+	binKE[binNo] += 0.5 * (angMom[0] * angMom[0] / I(0, 0) +
1476	+	angMom[1] * angMom[1] / I(1, 1) +
1477	+	angMom[2] * angMom[2] / I(2, 2));
1478	+	binDOF[binNo] += 3;
1479	+	}
1480	+	}
1481	+	}
1482	+
1483	+	#ifdef IS_MPI
1484	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binCount[0],
1485	+	nBins_, MPI::INT, MPI::SUM);
1486	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binMass[0],
1487	+	nBins_, MPI::REALTYPE, MPI::SUM);
1488	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binPx[0],
1489	+	nBins_, MPI::REALTYPE, MPI::SUM);
1490	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binPy[0],
1491	+	nBins_, MPI::REALTYPE, MPI::SUM);
1492	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binPz[0],
1493	+	nBins_, MPI::REALTYPE, MPI::SUM);
1494	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binKE[0],
1495	+	nBins_, MPI::REALTYPE, MPI::SUM);
1496	+	MPI::COMM_WORLD.Allreduce(MPI::IN_PLACE, &binDOF[0],
1497	+	nBins_, MPI::INT, MPI::SUM);
1498	+	#endif
1499	+
1500	+	Vector3d vel;
1501	+	RealType den;
1502	+	RealType temp;
1503	+	RealType z;
1504	+	for (int i = 0; i < nBins_; i++) {
1505	+	z = (((RealType)i + 0.5) / (RealType)nBins_) * hmat(2,2);
1506	+	vel.x() = binPx[i] / binMass[i];
1507	+	vel.y() = binPy[i] / binMass[i];
1508	+	vel.z() = binPz[i] / binMass[i];
1509	+
1510	+	den = binMass[i] * nBins_ * PhysicalConstants::densityConvert
1511	+	/ currentSnap_->getVolume() ;
1512	+
1513	+	if (binCount[i] > 0) {
1514	+	// only add values if there are things to add
1515	+	temp = 2.0 * binKE[i] / (binDOF[i] * PhysicalConstants::kb *
1516	+	PhysicalConstants::energyConvert);
1517	+
1518	+	for (unsigned int j = 0; j < outputMask_.size(); ++j) {
1519	+	if(outputMask_[j]) {
1520	+	switch(j) {
1521	+	case Z:
1522	+	dynamic_cast<Accumulator *>(data_[j].accumulator[i])->add(z);
1523	+	break;
1524	+	case TEMPERATURE:
1525	+	dynamic_cast<Accumulator *>(data_[j].accumulator[i])->add(temp);
1526	+	break;
1527	+	case VELOCITY:
1528	+	dynamic_cast<VectorAccumulator *>(data_[j].accumulator[i])->add(vel);
1529	+	break;
1530	+	case DENSITY:
1531	+	dynamic_cast<Accumulator *>(data_[j].accumulator[i])->add(den);
1532	+	break;
1533	+	}
1534	+	}
1535	+	}
1536	+	}
1537	+	}
1538	+	}
1539	+
1540	+	void RNEMD::getStarted() {
1541	+	if (!doRNEMD_) return;
1542	+	collectData();
1543	+	writeOutputFile();
1544	+	}
1545	+
1546	+	void RNEMD::parseOutputFileFormat(const std::string& format) {
1547	+	if (!doRNEMD_) return;
1548	+	StringTokenizer tokenizer(format, " ,;|\t\n\r");
1549	+
1550	+	while(tokenizer.hasMoreTokens()) {
1551	+	std::string token(tokenizer.nextToken());
1552	+	toUpper(token);
1553	+	OutputMapType::iterator i = outputMap_.find(token);
1554	+	if (i != outputMap_.end()) {
1555	+	outputMask_.set(i->second);
1556	+	} else {
1557	+	sprintf( painCave.errMsg,
1558	+	"RNEMD::parseOutputFileFormat: %s is not a recognized\n"
1559	+	"\toutputFileFormat keyword.\n", token.c_str() );
1560	+	painCave.isFatal = 0;
1561	+	painCave.severity = OPENMD_ERROR;
1562	+	simError();
1563	+	}
1564	+	}
1565	+	}
1566	+
1567	+	void RNEMD::writeOutputFile() {
1568	+	if (!doRNEMD_) return;
1569	+
1570	+	#ifdef IS_MPI
1571		// If we're the root node, should we print out the results
1572		int worldRank = MPI::COMM_WORLD.Get_rank();
1573		if (worldRank == 0) {
1574		#endif
1575	<
550	<	std::cout << time;
551	<	for (int j = 0; j < nBins_; j++)
552	<	std::cout << "\t" << valueHist[j] / (RealType)valueCount[j];
553	<	std::cout << "\n";
1575	>	rnemdFile_.open(rnemdFileName_.c_str(), std::ios::out | std::ios::trunc );
1576
1577	+	if( !rnemdFile_ ){
1578	+	sprintf( painCave.errMsg,
1579	+	"Could not open \"%s\" for RNEMD output.\n",
1580	+	rnemdFileName_.c_str());
1581	+	painCave.isFatal = 1;
1582	+	simError();
1583	+	}
1584	+
1585	+	Snapshot* currentSnap_ = info_->getSnapshotManager()->getCurrentSnapshot();
1586	+
1587	+	RealType time = currentSnap_->getTime();
1588	+	RealType avgArea;
1589	+	areaAccumulator_->getAverage(avgArea);
1590	+	RealType Jz = kineticExchange_ / (2.0 * time * avgArea)
1591	+	/ PhysicalConstants::energyConvert;
1592	+	Vector3d JzP = momentumExchange_ / (2.0 * time * avgArea);
1593	+
1594	+	rnemdFile_ << "#######################################################\n";
1595	+	rnemdFile_ << "# RNEMD {\n";
1596	+
1597	+	map<string, RNEMDMethod>::iterator mi;
1598	+	for(mi = stringToMethod_.begin(); mi != stringToMethod_.end(); ++mi) {
1599	+	if ( (*mi).second == rnemdMethod_)
1600	+	rnemdFile_ << "#    exchangeMethod  = \"" << (*mi).first << "\";\n";
1601	+	}
1602	+	map<string, RNEMDFluxType>::iterator fi;
1603	+	for(fi = stringToFluxType_.begin(); fi != stringToFluxType_.end(); ++fi) {
1604	+	if ( (*fi).second == rnemdFluxType_)
1605	+	rnemdFile_ << "#    fluxType  = \"" << (*fi).first << "\";\n";
1606	+	}
1607	+
1608	+	rnemdFile_ << "#    exchangeTime = " << exchangeTime_ << ";\n";
1609	+
1610	+	rnemdFile_ << "#    objectSelection = \""
1611	+	<< rnemdObjectSelection_ << "\";\n";
1612	+	rnemdFile_ << "#    slabWidth = " << slabWidth_ << ";\n";
1613	+	rnemdFile_ << "#    slabAcenter = " << slabACenter_ << ";\n";
1614	+	rnemdFile_ << "#    slabBcenter = " << slabBCenter_ << ";\n";
1615	+	rnemdFile_ << "# }\n";
1616	+	rnemdFile_ << "#######################################################\n";
1617	+	rnemdFile_ << "# RNEMD report:\n";
1618	+	rnemdFile_ << "#     running time = " << time << " fs\n";
1619	+	rnemdFile_ << "#     target flux:\n";
1620	+	rnemdFile_ << "#         kinetic = "
1621	+	<< kineticFlux_ / PhysicalConstants::energyConvert
1622	+	<< " (kcal/mol/A^2/fs)\n";
1623	+	rnemdFile_ << "#         momentum = " << momentumFluxVector_
1624	+	<< " (amu/A/fs^2)\n";
1625	+	rnemdFile_ << "#     target one-time exchanges:\n";
1626	+	rnemdFile_ << "#         kinetic = "
1627	+	<< kineticTarget_ / PhysicalConstants::energyConvert
1628	+	<< " (kcal/mol)\n";
1629	+	rnemdFile_ << "#         momentum = " << momentumTarget_
1630	+	<< " (amu*A/fs)\n";
1631	+	rnemdFile_ << "#     actual exchange totals:\n";
1632	+	rnemdFile_ << "#         kinetic = "
1633	+	<< kineticExchange_ / PhysicalConstants::energyConvert
1634	+	<< " (kcal/mol)\n";
1635	+	rnemdFile_ << "#         momentum = " << momentumExchange_
1636	+	<< " (amu*A/fs)\n";
1637	+	rnemdFile_ << "#     actual flux:\n";
1638	+	rnemdFile_ << "#         kinetic = " << Jz
1639	+	<< " (kcal/mol/A^2/fs)\n";
1640	+	rnemdFile_ << "#         momentum = " << JzP
1641	+	<< " (amu/A/fs^2)\n";
1642	+	rnemdFile_ << "#     exchange statistics:\n";
1643	+	rnemdFile_ << "#         attempted = " << trialCount_ << "\n";
1644	+	rnemdFile_ << "#         failed = " << failTrialCount_ << "\n";
1645	+	if (rnemdMethod_ == rnemdNIVS) {
1646	+	rnemdFile_ << "#         NIVS root-check errors = "
1647	+	<< failRootCount_ << "\n";
1648	+	}
1649	+	rnemdFile_ << "#######################################################\n";
1650	+
1651	+
1652	+
1653	+	//write title
1654	+	rnemdFile_ << "#";
1655	+	for (unsigned int i = 0; i < outputMask_.size(); ++i) {
1656	+	if (outputMask_[i]) {
1657	+	rnemdFile_ << "\t" << data_[i].title <<
1658	+	"(" << data_[i].units << ")";
1659	+	// add some extra tabs for column alignment
1660	+	if (data_[i].dataType == "Vector3d") rnemdFile_ << "\t\t";
1661	+	}
1662	+	}
1663	+	rnemdFile_ << std::endl;
1664	+
1665	+	rnemdFile_.precision(8);
1666	+
1667	+	for (int j = 0; j < nBins_; j++) {
1668	+
1669	+	for (unsigned int i = 0; i < outputMask_.size(); ++i) {
1670	+	if (outputMask_[i]) {
1671	+	if (data_[i].dataType == "RealType")
1672	+	writeReal(i,j);
1673	+	else if (data_[i].dataType == "Vector3d")
1674	+	writeVector(i,j);
1675	+	else {
1676	+	sprintf( painCave.errMsg,
1677	+	"RNEMD found an unknown data type for: %s ",
1678	+	data_[i].title.c_str());
1679	+	painCave.isFatal = 1;
1680	+	simError();
1681	+	}
1682	+	}
1683	+	}
1684	+	rnemdFile_ << std::endl;
1685	+
1686	+	}
1687	+
1688	+	rnemdFile_ << "#######################################################\n";
1689	+	rnemdFile_ << "# Standard Deviations in those quantities follow:\n";
1690	+	rnemdFile_ << "#######################################################\n";
1691	+
1692	+
1693	+	for (int j = 0; j < nBins_; j++) {
1694	+	rnemdFile_ << "#";
1695	+	for (unsigned int i = 0; i < outputMask_.size(); ++i) {
1696	+	if (outputMask_[i]) {
1697	+	if (data_[i].dataType == "RealType")
1698	+	writeRealStdDev(i,j);
1699	+	else if (data_[i].dataType == "Vector3d")
1700	+	writeVectorStdDev(i,j);
1701	+	else {
1702	+	sprintf( painCave.errMsg,
1703	+	"RNEMD found an unknown data type for: %s ",
1704	+	data_[i].title.c_str());
1705	+	painCave.isFatal = 1;
1706	+	simError();
1707	+	}
1708	+	}
1709	+	}
1710	+	rnemdFile_ << std::endl;
1711	+
1712	+	}
1713	+
1714	+	rnemdFile_.flush();
1715	+	rnemdFile_.close();
1716	+
1717		#ifdef IS_MPI
1718		}
1719		#endif
1720	+
1721		}
1722	+
1723	+	void RNEMD::writeReal(int index, unsigned int bin) {
1724	+	if (!doRNEMD_) return;
1725	+	assert(index >=0 && index < ENDINDEX);
1726	+	assert(int(bin) < nBins_);
1727	+	RealType s;
1728	+
1729	+	dynamic_cast<Accumulator *>(data_[index].accumulator[bin])->getAverage(s);
1730	+
1731	+	if (! isinf(s) && ! isnan(s)) {
1732	+	rnemdFile_ << "\t" << s;
1733	+	} else{
1734	+	sprintf( painCave.errMsg,
1735	+	"RNEMD detected a numerical error writing: %s for bin %d",
1736	+	data_[index].title.c_str(), bin);
1737	+	painCave.isFatal = 1;
1738	+	simError();
1739	+	}
1740	+	}
1741	+
1742	+	void RNEMD::writeVector(int index, unsigned int bin) {
1743	+	if (!doRNEMD_) return;
1744	+	assert(index >=0 && index < ENDINDEX);
1745	+	assert(int(bin) < nBins_);
1746	+	Vector3d s;
1747	+	dynamic_cast<VectorAccumulator*>(data_[index].accumulator[bin])->getAverage(s);
1748	+	if (isinf(s[0]) || isnan(s[0]) ||
1749	+	isinf(s[1]) || isnan(s[1]) ||
1750	+	isinf(s[2]) || isnan(s[2]) ) {
1751	+	sprintf( painCave.errMsg,
1752	+	"RNEMD detected a numerical error writing: %s for bin %d",
1753	+	data_[index].title.c_str(), bin);
1754	+	painCave.isFatal = 1;
1755	+	simError();
1756	+	} else {
1757	+	rnemdFile_ << "\t" << s[0] << "\t" << s[1] << "\t" << s[2];
1758	+	}
1759	+	}
1760	+
1761	+	void RNEMD::writeRealStdDev(int index, unsigned int bin) {
1762	+	if (!doRNEMD_) return;
1763	+	assert(index >=0 && index < ENDINDEX);
1764	+	assert(int(bin) < nBins_);
1765	+	RealType s;
1766	+
1767	+	dynamic_cast<Accumulator *>(data_[index].accumulator[bin])->getStdDev(s);
1768	+
1769	+	if (! isinf(s) && ! isnan(s)) {
1770	+	rnemdFile_ << "\t" << s;
1771	+	} else{
1772	+	sprintf( painCave.errMsg,
1773	+	"RNEMD detected a numerical error writing: %s std. dev. for bin %d",
1774	+	data_[index].title.c_str(), bin);
1775	+	painCave.isFatal = 1;
1776	+	simError();
1777	+	}
1778	+	}
1779	+
1780	+	void RNEMD::writeVectorStdDev(int index, unsigned int bin) {
1781	+	if (!doRNEMD_) return;
1782	+	assert(index >=0 && index < ENDINDEX);
1783	+	assert(int(bin) < nBins_);
1784	+	Vector3d s;
1785	+	dynamic_cast<VectorAccumulator*>(data_[index].accumulator[bin])->getStdDev(s);
1786	+	if (isinf(s[0]) || isnan(s[0]) ||
1787	+	isinf(s[1]) || isnan(s[1]) ||
1788	+	isinf(s[2]) || isnan(s[2]) ) {
1789	+	sprintf( painCave.errMsg,
1790	+	"RNEMD detected a numerical error writing: %s std. dev. for bin %d",
1791	+	data_[index].title.c_str(), bin);
1792	+	painCave.isFatal = 1;
1793	+	simError();
1794	+	} else {
1795	+	rnemdFile_ << "\t" << s[0] << "\t" << s[1] << "\t" << s[2];
1796	+	}
1797	+	}
1798		}
1799	+

Comparing:
trunk/src/integrators/RNEMD.cpp (property svn:keywords), Revision 1350 by gezelter, Thu May 21 18:56:45 2009 UTC vs.
branches/development/src/rnemd/RNEMD.cpp (property svn:keywords), Revision 1830 by gezelter, Wed Jan 9 22:02:30 2013 UTC

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

Diff Legend

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