HDK
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Groups Pages
LevelSetAdvect.h
Go to the documentation of this file.
1 // Copyright Contributors to the OpenVDB Project
2 // SPDX-License-Identifier: MPL-2.0
3 
4 /// @author Ken Museth
5 ///
6 /// @file tools/LevelSetAdvect.h
7 ///
8 /// @brief Hyperbolic advection of narrow-band level sets
9 
10 #ifndef OPENVDB_TOOLS_LEVEL_SET_ADVECT_HAS_BEEN_INCLUDED
11 #define OPENVDB_TOOLS_LEVEL_SET_ADVECT_HAS_BEEN_INCLUDED
12 
13 #include <tbb/parallel_for.h>
14 #include <tbb/parallel_reduce.h>
15 #include "LevelSetTracker.h"
16 #include "VelocityFields.h" // for EnrightField
17 #include <openvdb/Platform.h>
19 //#include <openvdb/util/CpuTimer.h>
20 #include <functional>
21 
22 
23 namespace openvdb {
25 namespace OPENVDB_VERSION_NAME {
26 namespace tools {
27 
28 /// @brief Hyperbolic advection of narrow-band level sets in an
29 /// external velocity field
30 ///
31 /// The @c FieldType template argument below refers to any functor
32 /// with the following interface (see tools/VelocityFields.h
33 /// for examples):
34 ///
35 /// @code
36 /// class VelocityField {
37 /// ...
38 /// public:
39 /// openvdb::VectorType operator() (const openvdb::Coord& xyz, ValueType time) const;
40 /// ...
41 /// };
42 /// @endcode
43 ///
44 /// @note The functor method returns the velocity field at coordinate
45 /// position xyz of the advection grid, and for the specified
46 /// time. Note that since the velocity is returned in the local
47 /// coordinate space of the grid that is being advected, the functor
48 /// typically depends on the transformation of that grid. This design
49 /// is chosen for performance reasons. Finally we will assume that the
50 /// functor method is NOT threadsafe (typically uses a ValueAccessor)
51 /// and that its lightweight enough that we can copy it per thread.
52 ///
53 /// The @c InterruptType template argument below refers to any class
54 /// with the following interface:
55 /// @code
56 /// class Interrupter {
57 /// ...
58 /// public:
59 /// void start(const char* name = nullptr) // called when computations begin
60 /// void end() // called when computations end
61 /// bool wasInterrupted(int percent=-1) // return true to break computation
62 ///};
63 /// @endcode
64 ///
65 /// @note If no template argument is provided for this InterruptType
66 /// the util::NullInterrupter is used which implies that all
67 /// interrupter calls are no-ops (i.e. incurs no computational overhead).
68 ///
69 
70 template<typename GridT,
71  typename FieldT = EnrightField<typename GridT::ValueType>,
72  typename InterruptT = util::NullInterrupter>
74 {
75 public:
76  using GridType = GridT;
78  using LeafRange = typename TrackerT::LeafRange;
79  using LeafType = typename TrackerT::LeafType;
80  using BufferType = typename TrackerT::BufferType;
81  using ValueType = typename TrackerT::ValueType;
82  using VectorType = typename FieldT::VectorType;
83 
84  /// Main constructor
85  LevelSetAdvection(GridT& grid, const FieldT& field, InterruptT* interrupt = nullptr):
86  mTracker(grid, interrupt), mField(field),
87  mSpatialScheme(math::HJWENO5_BIAS),
88  mTemporalScheme(math::TVD_RK2) {}
89 
90  virtual ~LevelSetAdvection() {}
91 
92  /// @brief Return the spatial finite difference scheme
93  math::BiasedGradientScheme getSpatialScheme() const { return mSpatialScheme; }
94  /// @brief Set the spatial finite difference scheme
95  void setSpatialScheme(math::BiasedGradientScheme scheme) { mSpatialScheme = scheme; }
96 
97  /// @brief Return the temporal integration scheme
98  math::TemporalIntegrationScheme getTemporalScheme() const { return mTemporalScheme; }
99  /// @brief Set the spatial finite difference scheme
100  void setTemporalScheme(math::TemporalIntegrationScheme scheme) { mTemporalScheme = scheme; }
101 
102  /// @brief Return the spatial finite difference scheme
104  return mTracker.getSpatialScheme();
105  }
106  /// @brief Set the spatial finite difference scheme
108  mTracker.setSpatialScheme(scheme);
109  }
110  /// @brief Return the temporal integration scheme
112  return mTracker.getTemporalScheme();
113  }
114  /// @brief Set the spatial finite difference scheme
116  mTracker.setTemporalScheme(scheme);
117  }
118 
119  /// @brief Return The number of normalizations performed per track or
120  /// normalize call.
121  int getNormCount() const { return mTracker.getNormCount(); }
122  /// @brief Set the number of normalizations performed per track or
123  /// normalize call.
124  void setNormCount(int n) { mTracker.setNormCount(n); }
125 
126  /// @brief Return the grain-size used for multi-threading
127  int getGrainSize() const { return mTracker.getGrainSize(); }
128  /// @brief Set the grain-size used for multi-threading.
129  /// @note A grain size of 0 or less disables multi-threading!
130  void setGrainSize(int grainsize) { mTracker.setGrainSize(grainsize); }
131 
132  /// Advect the level set from its current time, time0, to its
133  /// final time, time1. If time0>time1 backward advection is performed.
134  ///
135  /// @return number of CFL iterations used to advect from time0 to time1
136  size_t advect(ValueType time0, ValueType time1);
137 
138 private:
139  // disallow copy construction and copy by assignment!
140  LevelSetAdvection(const LevelSetAdvection&);// not implemented
141  LevelSetAdvection& operator=(const LevelSetAdvection&);// not implemented
142 
143  // This templated private struct implements all the level set magic.
144  template<typename MapT, math::BiasedGradientScheme SpatialScheme,
145  math::TemporalIntegrationScheme TemporalScheme>
146  struct Advect
147  {
148  /// Main constructor
149  Advect(LevelSetAdvection& parent);
150  /// Shallow copy constructor called by tbb::parallel_for() threads
151  Advect(const Advect& other);
152  /// Destructor
153  virtual ~Advect() { if (mIsMaster) this->clearField(); }
154  /// Advect the level set from its current time, time0, to its final time, time1.
155  /// @return number of CFL iterations
156  size_t advect(ValueType time0, ValueType time1);
157  /// Used internally by tbb::parallel_for()
158  void operator()(const LeafRange& r) const
159  {
160  if (mTask) mTask(const_cast<Advect*>(this), r);
161  else OPENVDB_THROW(ValueError, "task is undefined - don\'t call this method directly");
162  }
163  /// method calling tbb
164  void cook(const char* msg, size_t swapBuffer = 0);
165  /// Sample field and return the CFL time step
166  typename GridT::ValueType sampleField(ValueType time0, ValueType time1);
167  template <bool Aligned> void sample(const LeafRange& r, ValueType t0, ValueType t1);
168  inline void sampleXformed(const LeafRange& r, ValueType t0, ValueType t1)
169  {
170  this->sample<false>(r, t0, t1);
171  }
172  inline void sampleAligned(const LeafRange& r, ValueType t0, ValueType t1)
173  {
174  this->sample<true>(r, t0, t1);
175  }
176  void clearField();
177  // Convex combination of Phi and a forward Euler advection steps:
178  // Phi(result) = alpha * Phi(phi) + (1-alpha) * (Phi(0) - dt * Speed(speed)*|Grad[Phi(0)]|);
179  template <int Nominator, int Denominator>
180  void euler(const LeafRange&, ValueType, Index, Index);
181  inline void euler01(const LeafRange& r, ValueType t) {this->euler<0,1>(r, t, 0, 1);}
182  inline void euler12(const LeafRange& r, ValueType t) {this->euler<1,2>(r, t, 1, 1);}
183  inline void euler34(const LeafRange& r, ValueType t) {this->euler<3,4>(r, t, 1, 2);}
184  inline void euler13(const LeafRange& r, ValueType t) {this->euler<1,3>(r, t, 1, 2);}
185 
186  LevelSetAdvection& mParent;
187  VectorType* mVelocity;
188  size_t* mOffsets;
189  const MapT* mMap;
190  typename std::function<void (Advect*, const LeafRange&)> mTask;
191  const bool mIsMaster;
192  }; // end of private Advect struct
193 
194  template<math::BiasedGradientScheme SpatialScheme>
195  size_t advect1(ValueType time0, ValueType time1);
196 
197  template<math::BiasedGradientScheme SpatialScheme,
198  math::TemporalIntegrationScheme TemporalScheme>
199  size_t advect2(ValueType time0, ValueType time1);
200 
201  template<math::BiasedGradientScheme SpatialScheme,
202  math::TemporalIntegrationScheme TemporalScheme,
203  typename MapType>
204  size_t advect3(ValueType time0, ValueType time1);
205 
206  TrackerT mTracker;
207  //each thread needs a deep copy of the field since it might contain a ValueAccessor
208  const FieldT mField;
209  math::BiasedGradientScheme mSpatialScheme;
210  math::TemporalIntegrationScheme mTemporalScheme;
211 
212 };//end of LevelSetAdvection
213 
214 
215 template<typename GridT, typename FieldT, typename InterruptT>
216 size_t
218 {
219  switch (mSpatialScheme) {
220  case math::FIRST_BIAS:
221  return this->advect1<math::FIRST_BIAS >(time0, time1);
222  case math::SECOND_BIAS:
223  return this->advect1<math::SECOND_BIAS >(time0, time1);
224  case math::THIRD_BIAS:
225  return this->advect1<math::THIRD_BIAS >(time0, time1);
226  case math::WENO5_BIAS:
227  return this->advect1<math::WENO5_BIAS >(time0, time1);
228  case math::HJWENO5_BIAS:
229  return this->advect1<math::HJWENO5_BIAS>(time0, time1);
230  default:
231  OPENVDB_THROW(ValueError, "Spatial difference scheme not supported!");
232  }
233  return 0;
234 }
235 
236 
237 template<typename GridT, typename FieldT, typename InterruptT>
238 template<math::BiasedGradientScheme SpatialScheme>
239 size_t
241 {
242  switch (mTemporalScheme) {
243  case math::TVD_RK1:
244  return this->advect2<SpatialScheme, math::TVD_RK1>(time0, time1);
245  case math::TVD_RK2:
246  return this->advect2<SpatialScheme, math::TVD_RK2>(time0, time1);
247  case math::TVD_RK3:
248  return this->advect2<SpatialScheme, math::TVD_RK3>(time0, time1);
249  default:
250  OPENVDB_THROW(ValueError, "Temporal integration scheme not supported!");
251  }
252  return 0;
253 }
254 
255 
256 template<typename GridT, typename FieldT, typename InterruptT>
257 template<math::BiasedGradientScheme SpatialScheme, math::TemporalIntegrationScheme TemporalScheme>
258 size_t
259 LevelSetAdvection<GridT, FieldT, InterruptT>::advect2(ValueType time0, ValueType time1)
260 {
261  const math::Transform& trans = mTracker.grid().transform();
262  if (trans.mapType() == math::UniformScaleMap::mapType()) {
263  return this->advect3<SpatialScheme, TemporalScheme, math::UniformScaleMap>(time0, time1);
264  } else if (trans.mapType() == math::UniformScaleTranslateMap::mapType()) {
265  return this->advect3<SpatialScheme, TemporalScheme, math::UniformScaleTranslateMap>(
266  time0, time1);
267  } else if (trans.mapType() == math::UnitaryMap::mapType()) {
268  return this->advect3<SpatialScheme, TemporalScheme, math::UnitaryMap >(time0, time1);
269  } else if (trans.mapType() == math::TranslationMap::mapType()) {
270  return this->advect3<SpatialScheme, TemporalScheme, math::TranslationMap>(time0, time1);
271  } else {
272  OPENVDB_THROW(ValueError, "MapType not supported!");
273  }
274  return 0;
275 }
276 
277 
278 template<typename GridT, typename FieldT, typename InterruptT>
279 template<
280  math::BiasedGradientScheme SpatialScheme,
281  math::TemporalIntegrationScheme TemporalScheme,
282  typename MapT>
283 size_t
284 LevelSetAdvection<GridT, FieldT, InterruptT>::advect3(ValueType time0, ValueType time1)
285 {
286  Advect<MapT, SpatialScheme, TemporalScheme> tmp(*this);
287  return tmp.advect(time0, time1);
288 }
289 
290 
291 ///////////////////////////////////////////////////////////////////////
292 
293 
294 template<typename GridT, typename FieldT, typename InterruptT>
295 template<
296  typename MapT,
297  math::BiasedGradientScheme SpatialScheme,
298  math::TemporalIntegrationScheme TemporalScheme>
299 inline
300 LevelSetAdvection<GridT, FieldT, InterruptT>::
301 Advect<MapT, SpatialScheme, TemporalScheme>::
302 Advect(LevelSetAdvection& parent)
303  : mParent(parent)
304  , mVelocity(nullptr)
305  , mOffsets(nullptr)
306  , mMap(parent.mTracker.grid().transform().template constMap<MapT>().get())
307  , mTask(0)
308  , mIsMaster(true)
309 {
310 }
311 
312 
313 template<typename GridT, typename FieldT, typename InterruptT>
314 template<
315  typename MapT,
316  math::BiasedGradientScheme SpatialScheme,
317  math::TemporalIntegrationScheme TemporalScheme>
318 inline
319 LevelSetAdvection<GridT, FieldT, InterruptT>::
320 Advect<MapT, SpatialScheme, TemporalScheme>::
321 Advect(const Advect& other)
322  : mParent(other.mParent)
323  , mVelocity(other.mVelocity)
324  , mOffsets(other.mOffsets)
325  , mMap(other.mMap)
326  , mTask(other.mTask)
327  , mIsMaster(false)
328 {
329 }
330 
331 
332 template<typename GridT, typename FieldT, typename InterruptT>
333 template<
334  typename MapT,
335  math::BiasedGradientScheme SpatialScheme,
336  math::TemporalIntegrationScheme TemporalScheme>
337 inline size_t
340 advect(ValueType time0, ValueType time1)
341 {
342  namespace ph = std::placeholders;
343 
344  //util::CpuTimer timer;
345  size_t countCFL = 0;
346  if ( math::isZero(time0 - time1) ) return countCFL;
347  const bool isForward = time0 < time1;
348  while ((isForward ? time0<time1 : time0>time1) && mParent.mTracker.checkInterrupter()) {
349  /// Make sure we have enough temporal auxiliary buffers
350  //timer.start( "\nallocate buffers" );
351  mParent.mTracker.leafs().rebuildAuxBuffers(TemporalScheme == math::TVD_RK3 ? 2 : 1);
352  //timer.stop();
353 
354  const ValueType dt = this->sampleField(time0, time1);
355  if ( math::isZero(dt) ) break;//V is essentially zero so terminate
356 
357  OPENVDB_NO_UNREACHABLE_CODE_WARNING_BEGIN //switch is resolved at compile-time
358  switch(TemporalScheme) {
359  case math::TVD_RK1:
360  // Perform one explicit Euler step: t1 = t0 + dt
361  // Phi_t1(1) = Phi_t0(0) - dt * VdotG_t0(0)
362  mTask = std::bind(&Advect::euler01, ph::_1, ph::_2, dt);
363 
364  // Cook and swap buffer 0 and 1 such that Phi_t1(0) and Phi_t0(1)
365  this->cook("Advecting level set using TVD_RK1", 1);
366  break;
367  case math::TVD_RK2:
368  // Perform one explicit Euler step: t1 = t0 + dt
369  // Phi_t1(1) = Phi_t0(0) - dt * VdotG_t0(0)
370  mTask = std::bind(&Advect::euler01, ph::_1, ph::_2, dt);
371 
372  // Cook and swap buffer 0 and 1 such that Phi_t1(0) and Phi_t0(1)
373  this->cook("Advecting level set using TVD_RK1 (step 1 of 2)", 1);
374 
375  // Convex combine explict Euler step: t2 = t0 + dt
376  // Phi_t2(1) = 1/2 * Phi_t0(1) + 1/2 * (Phi_t1(0) - dt * V.Grad_t1(0))
377  mTask = std::bind(&Advect::euler12, ph::_1, ph::_2, dt);
378 
379  // Cook and swap buffer 0 and 1 such that Phi_t2(0) and Phi_t1(1)
380  this->cook("Advecting level set using TVD_RK1 (step 2 of 2)", 1);
381  break;
382  case math::TVD_RK3:
383  // Perform one explicit Euler step: t1 = t0 + dt
384  // Phi_t1(1) = Phi_t0(0) - dt * VdotG_t0(0)
385  mTask = std::bind(&Advect::euler01, ph::_1, ph::_2, dt);
386 
387  // Cook and swap buffer 0 and 1 such that Phi_t1(0) and Phi_t0(1)
388  this->cook("Advecting level set using TVD_RK3 (step 1 of 3)", 1);
389 
390  // Convex combine explict Euler step: t2 = t0 + dt/2
391  // Phi_t2(2) = 3/4 * Phi_t0(1) + 1/4 * (Phi_t1(0) - dt * V.Grad_t1(0))
392  mTask = std::bind(&Advect::euler34, ph::_1, ph::_2, dt);
393 
394  // Cook and swap buffer 0 and 2 such that Phi_t2(0) and Phi_t1(2)
395  this->cook("Advecting level set using TVD_RK3 (step 2 of 3)", 2);
396 
397  // Convex combine explict Euler step: t3 = t0 + dt
398  // Phi_t3(2) = 1/3 * Phi_t0(1) + 2/3 * (Phi_t2(0) - dt * V.Grad_t2(0)
399  mTask = std::bind(&Advect::euler13, ph::_1, ph::_2, dt);
400 
401  // Cook and swap buffer 0 and 2 such that Phi_t3(0) and Phi_t2(2)
402  this->cook("Advecting level set using TVD_RK3 (step 3 of 3)", 2);
403  break;
404  default:
405  OPENVDB_THROW(ValueError, "Temporal integration scheme not supported!");
406  }//end of compile-time resolved switch
408 
409  time0 += isForward ? dt : -dt;
410  ++countCFL;
411  mParent.mTracker.leafs().removeAuxBuffers();
412  this->clearField();
413  /// Track the narrow band
414  mParent.mTracker.track();
415  }//end wile-loop over time
416  return countCFL;//number of CLF propagation steps
417 }
418 
419 
420 template<typename GridT, typename FieldT, typename InterruptT>
421 template<
422  typename MapT,
423  math::BiasedGradientScheme SpatialScheme,
424  math::TemporalIntegrationScheme TemporalScheme>
425 inline typename GridT::ValueType
426 LevelSetAdvection<GridT, FieldT, InterruptT>::
427 Advect<MapT, SpatialScheme, TemporalScheme>::
428 sampleField(ValueType time0, ValueType time1)
429 {
430  namespace ph = std::placeholders;
431 
432  const int grainSize = mParent.mTracker.getGrainSize();
433  const size_t leafCount = mParent.mTracker.leafs().leafCount();
434  if (leafCount==0) return ValueType(0.0);
435 
436  // Compute the prefix sum of offsets to active voxels
437  size_t size=0, voxelCount=mParent.mTracker.leafs().getPrefixSum(mOffsets, size, grainSize);
438 
439  // Sample the velocity field
440  if (mParent.mField.transform() == mParent.mTracker.grid().transform()) {
441  mTask = std::bind(&Advect::sampleAligned, ph::_1, ph::_2, time0, time1);
442  } else {
443  mTask = std::bind(&Advect::sampleXformed, ph::_1, ph::_2, time0, time1);
444  }
445  assert(voxelCount == mParent.mTracker.grid().activeVoxelCount());
446  mVelocity = new VectorType[ voxelCount ];
447  this->cook("Sampling advection field");
448 
449  // Find the extrema of the magnitude of the velocities
450  ValueType maxAbsV = 0;
451  VectorType* v = mVelocity;
452  for (size_t i = 0; i < voxelCount; ++i, ++v) {
453  maxAbsV = math::Max(maxAbsV, ValueType(v->lengthSqr()));
454  }
455 
456  // Compute the CFL number
458  static const ValueType CFL = (TemporalScheme == math::TVD_RK1 ? ValueType(0.3) :
459  TemporalScheme == math::TVD_RK2 ? ValueType(0.9) :
460  ValueType(1.0))/math::Sqrt(ValueType(3.0));
461  const ValueType dt = math::Abs(time1 - time0), dx = mParent.mTracker.voxelSize();
462  return math::Min(dt, ValueType(CFL*dx/math::Sqrt(maxAbsV)));
463 }
464 
465 
466 template<typename GridT, typename FieldT, typename InterruptT>
467 template<
468  typename MapT,
469  math::BiasedGradientScheme SpatialScheme,
470  math::TemporalIntegrationScheme TemporalScheme>
471 template<bool Aligned>
472 inline void
473 LevelSetAdvection<GridT, FieldT, InterruptT>::
474 Advect<MapT, SpatialScheme, TemporalScheme>::
475 sample(const LeafRange& range, ValueType time0, ValueType time1)
476 {
477  const bool isForward = time0 < time1;
478  using VoxelIterT = typename LeafType::ValueOnCIter;
479  const MapT& map = *mMap;
480  const FieldT field( mParent.mField );
481  mParent.mTracker.checkInterrupter();
482  for (typename LeafRange::Iterator leafIter = range.begin(); leafIter; ++leafIter) {
483  VectorType* vel = mVelocity + mOffsets[ leafIter.pos() ];
484  for (VoxelIterT iter = leafIter->cbeginValueOn(); iter; ++iter, ++vel) {
486  const VectorType v = Aligned ? field(iter.getCoord(), time0) ://resolved at compile time
487  field(map.applyMap(iter.getCoord().asVec3d()), time0);
488  *vel = isForward ? v : -v;
490  }
491  }
492 }
493 
494 
495 template<typename GridT, typename FieldT, typename InterruptT>
496 template<
497  typename MapT,
498  math::BiasedGradientScheme SpatialScheme,
499  math::TemporalIntegrationScheme TemporalScheme>
500 inline void
501 LevelSetAdvection<GridT, FieldT, InterruptT>::
502 Advect<MapT, SpatialScheme, TemporalScheme>::
503 clearField()
504 {
505  delete [] mOffsets;
506  delete [] mVelocity;
507  mOffsets = nullptr;
508  mVelocity = nullptr;
509 }
510 
511 
512 template<typename GridT, typename FieldT, typename InterruptT>
513 template<
514  typename MapT,
515  math::BiasedGradientScheme SpatialScheme,
516  math::TemporalIntegrationScheme TemporalScheme>
517 inline void
518 LevelSetAdvection<GridT, FieldT, InterruptT>::
519 Advect<MapT, SpatialScheme, TemporalScheme>::
520 cook(const char* msg, size_t swapBuffer)
521 {
522  mParent.mTracker.startInterrupter( msg );
523 
524  const int grainSize = mParent.mTracker.getGrainSize();
525  const LeafRange range = mParent.mTracker.leafs().leafRange(grainSize);
526 
527  grainSize == 0 ? (*this)(range) : tbb::parallel_for(range, *this);
528 
529  mParent.mTracker.leafs().swapLeafBuffer(swapBuffer, grainSize == 0);
530 
531  mParent.mTracker.endInterrupter();
532 }
533 
534 
535 // Convex combination of Phi and a forward Euler advection steps:
536 // Phi(result) = alpha * Phi(phi) + (1-alpha) * (Phi(0) - dt * V.Grad(0));
537 template<typename GridT, typename FieldT, typename InterruptT>
538 template<
539  typename MapT,
540  math::BiasedGradientScheme SpatialScheme,
541  math::TemporalIntegrationScheme TemporalScheme>
542 template <int Nominator, int Denominator>
543 inline void
544 LevelSetAdvection<GridT, FieldT, InterruptT>::
545 Advect<MapT, SpatialScheme, TemporalScheme>::
546 euler(const LeafRange& range, ValueType dt, Index phiBuffer, Index resultBuffer)
547 {
548  using SchemeT = math::BIAS_SCHEME<SpatialScheme>;
549  using StencilT = typename SchemeT::template ISStencil<GridType>::StencilType;
550  using VoxelIterT = typename LeafType::ValueOnCIter;
551  using GradT = math::GradientBiased<MapT, SpatialScheme>;
552 
553  static const ValueType Alpha = ValueType(Nominator)/ValueType(Denominator);
554  static const ValueType Beta = ValueType(1) - Alpha;
555 
556  mParent.mTracker.checkInterrupter();
557  const MapT& map = *mMap;
558  StencilT stencil(mParent.mTracker.grid());
559  for (typename LeafRange::Iterator leafIter = range.begin(); leafIter; ++leafIter) {
560  const VectorType* vel = mVelocity + mOffsets[ leafIter.pos() ];
561  const ValueType* phi = leafIter.buffer(phiBuffer).data();
562  ValueType* result = leafIter.buffer(resultBuffer).data();
563  for (VoxelIterT voxelIter = leafIter->cbeginValueOn(); voxelIter; ++voxelIter, ++vel) {
564  const Index i = voxelIter.pos();
565  stencil.moveTo(voxelIter);
566  const ValueType a =
567  stencil.getValue() - dt * vel->dot(GradT::result(map, stencil, *vel));
568  result[i] = Nominator ? Alpha * phi[i] + Beta * a : a;
569  }//loop over active voxels in the leaf of the mask
570  }//loop over leafs of the level set
571 }
572 
573 
574 ////////////////////////////////////////
575 
576 
577 // Explicit Template Instantiation
578 
579 #ifdef OPENVDB_USE_EXPLICIT_INSTANTIATION
580 
581 #ifdef OPENVDB_INSTANTIATE_LEVELSETADVECT
583 #endif
584 
585 OPENVDB_INSTANTIATE_CLASS LevelSetAdvection<FloatGrid, DiscreteField<Vec3SGrid, BoxSampler>, util::NullInterrupter>;
586 OPENVDB_INSTANTIATE_CLASS LevelSetAdvection<DoubleGrid, DiscreteField<Vec3SGrid, BoxSampler>, util::NullInterrupter>;
587 
588 #endif // OPENVDB_USE_EXPLICIT_INSTANTIATION
589 
590 
591 } // namespace tools
592 } // namespace OPENVDB_VERSION_NAME
593 } // namespace openvdb
594 
595 #endif // OPENVDB_TOOLS_LEVEL_SET_ADVECT_HAS_BEEN_INCLUDED
#define OPENVDB_NO_TYPE_CONVERSION_WARNING_END
Definition: Platform.h:228
void parallel_for(int64_t start, int64_t end, std::function< void(int64_t index)> &&task, parallel_options opt=parallel_options(0, Split_Y, 1))
Definition: parallel.h:127
TemporalIntegrationScheme
Temporal integration schemes.
GLenum GLint * range
Definition: glcorearb.h:1925
LevelSetAdvection(GridT &grid, const FieldT &field, InterruptT *interrupt=nullptr)
Main constructor.
size_t advect(ValueType time0, ValueType time1)
void setTemporalScheme(math::TemporalIntegrationScheme s)
Set the spatial finite difference scheme.
const GLdouble * v
Definition: glcorearb.h:837
GLsizei const GLfloat * value
Definition: glcorearb.h:824
void setGrainSize(int grainsize)
Set the grain-size used for multi-threading.
GLboolean GLboolean GLboolean GLboolean a
Definition: glcorearb.h:1222
#define OPENVDB_USE_VERSION_NAMESPACE
Definition: version.h:239
**But if you need a result
Definition: thread.h:613
Performs multi-threaded interface tracking of narrow band level sets.
math::TemporalIntegrationScheme getTemporalScheme() const
BiasedGradientScheme
Biased Gradients are limited to non-centered differences.
void setNormCount(int n)
Set the number of normalizations performed per track or normalize call.
GLdouble n
Definition: glcorearb.h:2008
#define OPENVDB_INSTANTIATE_CLASS
math::BiasedGradientScheme getSpatialScheme() const
Coord Abs(const Coord &xyz)
Definition: Coord.h:517
void setSpatialScheme(math::BiasedGradientScheme scheme)
Set the spatial finite difference scheme.
GA_API const UT_StringHolder trans
int getNormCount() const
Return The number of normalizations performed per track or normalize call.
float Sqrt(float x)
Return the square root of a floating-point value.
Definition: Math.h:761
void setTrackerSpatialScheme(math::BiasedGradientScheme scheme)
Set the spatial finite difference scheme.
auto get(const UT_ARTIterator< T > &it) -> decltype(it.key())
Definition: UT_ARTMap.h:1073
void setNormCount(int n)
Set the number of normalizations performed per track or normalize call.
void setSpatialScheme(math::BiasedGradientScheme s)
Set the spatial finite difference scheme.
Performs multi-threaded interface tracking of narrow band level sets. This is the building-block for ...
bool isApproxZero(const Type &x)
Return true if x is equal to zero to within the default floating-point comparison tolerance...
Definition: Math.h:349
GA_API const UT_StringHolder transform
#define OPENVDB_NO_UNREACHABLE_CODE_WARNING_BEGIN
Definition: Platform.h:140
GLdouble t
Definition: glad.h:2397
math::BiasedGradientScheme getSpatialScheme() const
Return the spatial finite difference scheme.
void setGrainSize(int grainsize)
Set the grain-size used for multi-threading.
math::TemporalIntegrationScheme getTrackerTemporalScheme() const
Return the temporal integration scheme.
static Name mapType()
Return UnitaryMap.
Definition: Maps.h:1717
GLsizeiptr size
Definition: glcorearb.h:664
Defines two simple wrapper classes for advection velocity fields as well as VelocitySampler and Veloc...
#define OPENVDB_NO_TYPE_CONVERSION_WARNING_BEGIN
Bracket code with OPENVDB_NO_TYPE_CONVERSION_WARNING_BEGIN/_END, to inhibit warnings about type conve...
Definition: Platform.h:227
typename LeafManagerType::BufferType BufferType
math::BiasedGradientScheme getTrackerSpatialScheme() const
Return the spatial finite difference scheme.
const Type & Min(const Type &a, const Type &b)
Return the minimum of two values.
Definition: Math.h:656
#define OPENVDB_NO_UNREACHABLE_CODE_WARNING_END
Definition: Platform.h:141
typename LeafManagerType::LeafRange LeafRange
GLboolean r
Definition: glcorearb.h:1222
GLint GLfloat GLint stencil
Definition: glcorearb.h:1278
int getGrainSize() const
Return the grain-size used for multi-threading.
__hostdev__ float Sqrt(float x)
Return the square root of a floating-point value.
Definition: NanoVDB.h:1214
GA_API const UT_StringHolder Alpha
#define OPENVDB_VERSION_NAME
The version namespace name for this library version.
Definition: version.h:119
bool ValueType
Definition: NanoVDB.h:5729
Hyperbolic advection of narrow-band level sets in an external velocity field.
const Type & Max(const Type &a, const Type &b)
Return the maximum of two values.
Definition: Math.h:595
math::TemporalIntegrationScheme getTemporalScheme() const
Return the temporal integration scheme.
LevelSetTracker< GridT, InterruptT > TrackerT
bool isZero(const Type &x)
Return true if x is exactly equal to zero.
Definition: Math.h:337
#define OPENVDB_THROW(exception, message)
Definition: Exceptions.h:74
void setTrackerTemporalScheme(math::TemporalIntegrationScheme scheme)
Set the spatial finite difference scheme.
void setTemporalScheme(math::TemporalIntegrationScheme scheme)
Set the spatial finite difference scheme.