openvdb::v12_0::points Namespace Reference

Namespaces
	attribute_traits
	future
	A container for ABI=5 to help ease introduction of upcoming features.
	index
	internal
	point_attribute_internal
	point_move_internal
	point_rasterize_internal
	transfer_internal

Classes
struct	AccessorEval
	Accessor to call unsafe get and set methods based on templated Codec and Value. More...
struct	AccessorEval< UnknownCodec, ValueType >
	Partial specialization when Codec is not known at compile-time to use the supplied functor instead. More...
class	AttributeArray
	Base class for storing attribute data. More...
class	AttributeHandle
class	AttributeHashFilter
class	AttributeSet
	Ordered collection of uniquely-named attribute arrays. More...
class	AttributeWriteHandle
	Write-able version of AttributeHandle. More...
class	BBoxFilter
class	BinaryFilter
class	CachedDeformer
	A Deformer that caches the resulting positions from evaluating another Deformer. More...
struct	DeformerTraits
	Deformer Traits for optionally configuring deformers to be applied in index-space. The default is world-space. More...
struct	DummySampleType
struct	FilterTraits
struct	FilterTraits< BBoxFilter >
struct	FilterTraits< BinaryFilter< T0, T1, And > >
struct	FilterTraits< LevelSetFilter< T > >
struct	FixedPointCodec
class	FrustumRasterizer
	Efficient rasterization of one or more VDB Points grids into a linear or frustum volume with the option to bake in camera or geometry motion blur. More...
struct	FrustumRasterizerMask
struct	FrustumRasterizerSettings
	A group of shared settings to be used in the Volume Rasterizer. More...
struct	GroupCodec
class	GroupFilter
	Index filtering on group membership. More...
class	GroupHandle
class	GroupWriteHandle
class	IndexIter
	A forward iterator over array indices with filtering IteratorT can be either IndexIter or ValueIndexIter (or some custom index iterator) FilterT should be a struct or class with a valid() method than can be evaluated per index Here's a simple filter example that only accepts even indices: More...
class	LevelSetFilter
class	MultiGroupFilter
struct	NullCodec
struct	NullDeformer
	No-op deformer (adheres to the deformer interface documented in PointMove.h) More...
class	NullFilter
	A no-op filter that can be used when iterating over all indices. More...
class	PointAttributeVector
	Point-partitioner compatible STL vector attribute wrapper for convenience. More...
class	PointDataLeafNode
struct	PositionRange
class	RandomLeafFilter
class	RasterCamera
	A camera class that provides an interface for camera motion blur when rasterizing. More...
struct	RasterGroups
	A struct that stores all include/exclude attribute names as strings and is internally converted into the resolved MultiGroupFilter. More...
struct	SampleWithRounding
class	StringAttributeHandle
class	StringAttributeWriteHandle
struct	StringCodec
class	StringMetaCache
	Class to compute a string->index map from all string:N metadata. More...
class	StringMetaInserter
	Class to help with insertion of keyed string values into metadata. More...
struct	TransformTransfer
	The TransformTransfer module should be used if the source transform of the input points and the target transforms of the destination volumes differ. The default rasterizer will skip index to world (and vice versa) transformations unless a transfer scheme derives from a TransformTransfer. More...
struct	TreeConverter
	Similiar to ValueConverter, but allows for tree configuration conversion to a PointDataTree. ValueConverter<PointDataIndex32> cannot be used as a PointDataLeafNode is not a specialization of LeafNode. More...
struct	TrilinearTraits
struct	TrilinearTraits< ValueT, false >
struct	TruncateCodec
class	TypedAttributeArray
	Typed class for storing attribute data. More...
struct	UnitRange
struct	UnitVecCodec
struct	UnknownCodec
class	ValueMaskFilter
	Index filtering on active / inactive state of host voxel. More...
class	ValueVoxelCIter
	A forward iterator over array indices in a single voxel. More...
struct	VolumeTransfer
	The VolumeTransfer module provides methods to automatically setup and access destination buffers for multiple target volumes of arbitrary types. Deriving from a VolumeTransfer ensures that the available buffers correlate to the order of the provided tree arguments. More...
struct	VolumeTransfer< TreeT >
	VolumeTransfer specialization for a single target volume. More...

Typedefs
using	StringAttributeArray = TypedAttributeArray< Index, StringCodec< false >>
using	GroupAttributeArray = TypedAttributeArray< GroupType, GroupCodec >
using	GroupType = uint8_t
using	ActiveFilter = ValueMaskFilter< true >
using	InactiveFilter = ValueMaskFilter< false >
using	PointDataTree = tree::Tree< tree::RootNode< tree::InternalNode< tree::InternalNode< PointDataLeafNode< PointDataIndex32, 3 >, 4 >, 5 >>>
	Point index tree configured to match the default VDB configurations. More...
using	PointDataGrid = Grid< PointDataTree >
	Point data grid. More...

Enumerations
enum	RasterMode { ACCUMULATE = 0, MAXIMUM, AVERAGE }
	How to composite points into a volume. More...

Functions
template<typename IntegerT , typename FloatT >
IntegerT	floatingPointToFixedPoint (const FloatT s)
template<typename FloatT , typename IntegerT >
FloatT	fixedPointToFloatingPoint (const IntegerT s)
template<typename IntegerVectorT , typename FloatT >
IntegerVectorT	floatingPointToFixedPoint (const math::Vec3< FloatT > &v)
template<typename FloatVectorT , typename IntegerT >
FloatVectorT	fixedPointToFloatingPoint (const math::Vec3< IntegerT > &v)
bool	isString (const AttributeArray &array)
bool	isGroup (const AttributeArray &array)
template<typename IterT >
Index64	iterCount (const IterT &iter)
	Count up the number of times the iterator can iterate. More...
template<typename PointDataGridT , typename VelGridT , typename AdvectFilterT = NullFilter, typename FilterT = NullFilter>
void	advectPoints (PointDataGridT &points, const VelGridT &velocity, const Index integrationOrder, const double dt, const Index timeSteps, const AdvectFilterT &advectFilter=NullFilter(), const FilterT &filter=NullFilter(), const bool cached=true)
	Advect points in a PointDataGrid through a velocity grid. More...
template<typename PointDataTreeT >
void	appendAttribute (PointDataTreeT &tree, const Name &name, const NamePair &type, const Index strideOrTotalSize=1, const bool constantStride=true, const Metadata *defaultValue=nullptr, const bool hidden=false, const bool transient=false)
	Appends a new attribute to the VDB tree (this method does not require a templated AttributeType) More...
template<typename ValueType , typename CodecType = NullCodec, typename PointDataTreeT >
void	appendAttribute (PointDataTreeT &tree, const std::string &name, const ValueType &uniformValue=point_attribute_internal::Default< ValueType >::value(), const Index strideOrTotalSize=1, const bool constantStride=true, const TypedMetadata< ValueType > *defaultValue=nullptr, const bool hidden=false, const bool transient=false)
	Appends a new attribute to the VDB tree. More...
template<typename ValueType , typename PointDataTreeT >
void	collapseAttribute (PointDataTreeT &tree, const Name &name, const ValueType &uniformValue=point_attribute_internal::Default< ValueType >::value())
	Collapse the attribute into a uniform value. More...
template<typename PointDataTreeT >
void	dropAttributes (PointDataTreeT &tree, const std::vector< size_t > &indices)
	Drops attributes from the VDB tree. More...
template<typename PointDataTreeT >
void	dropAttributes (PointDataTreeT &tree, const std::vector< Name > &names)
	Drops attributes from the VDB tree. More...
template<typename PointDataTreeT >
void	dropAttribute (PointDataTreeT &tree, const size_t &index)
	Drop one attribute from the VDB tree (convenience method). More...
template<typename PointDataTreeT >
void	dropAttribute (PointDataTreeT &tree, const Name &name)
	Drop one attribute from the VDB tree (convenience method). More...
template<typename PointDataTreeT >
void	renameAttributes (PointDataTreeT &tree, const std::vector< Name > &oldNames, const std::vector< Name > &newNames)
	Rename attributes in a VDB tree. More...
template<typename PointDataTreeT >
void	renameAttribute (PointDataTreeT &tree, const Name &oldName, const Name &newName)
	Rename an attribute in a VDB tree. More...
template<typename PointDataTreeT >
void	compactAttributes (PointDataTreeT &tree)
	Compact attributes in a VDB tree (if possible). More...
template<typename CompressionT , typename PointDataGridT , typename PositionArrayT , typename PointIndexGridT >
PointDataGridT::Ptr	createPointDataGrid (const PointIndexGridT &pointIndexGrid, const PositionArrayT &positions, const math::Transform &xform, const Metadata *positionDefaultValue=nullptr)
	Localises points with position into a PointDataGrid into two stages: allocation of the leaf attribute data and population of the positions. More...
template<typename CompressionT , typename PointDataGridT , typename ValueT >
PointDataGridT::Ptr	createPointDataGrid (const std::vector< ValueT > &positions, const math::Transform &xform, const Metadata *positionDefaultValue=nullptr)
	Convenience method to create a PointDataGrid from a std::vector of point positions. More...
template<typename PointDataTreeT , typename PointIndexTreeT , typename PointArrayT >
void	populateAttribute (PointDataTreeT &tree, const PointIndexTreeT &pointIndexTree, const openvdb::Name &attributeName, const PointArrayT &data, const Index stride=1, const bool insertMetadata=true)
	Stores point attribute data in an existing PointDataGrid attribute. More...
template<typename PositionAttribute , typename PointDataGridT , typename FilterT = NullFilter>
void	convertPointDataGridPosition (PositionAttribute &positionAttribute, const PointDataGridT &grid, const std::vector< Index64 > &pointOffsets, const Index64 startOffset, const FilterT &filter=NullFilter(), const bool inCoreOnly=false)
	Convert the position attribute from a Point Data Grid. More...
template<typename TypedAttribute , typename PointDataTreeT , typename FilterT = NullFilter>
void	convertPointDataGridAttribute (TypedAttribute &attribute, const PointDataTreeT &tree, const std::vector< Index64 > &pointOffsets, const Index64 startOffset, const unsigned arrayIndex, const Index stride=1, const FilterT &filter=NullFilter(), const bool inCoreOnly=false)
	Convert the attribute from a PointDataGrid. More...
template<typename Group , typename PointDataTreeT , typename FilterT = NullFilter>
void	convertPointDataGridGroup (Group &group, const PointDataTreeT &tree, const std::vector< Index64 > &pointOffsets, const Index64 startOffset, const AttributeSet::Descriptor::GroupIndex index, const FilterT &filter=NullFilter(), const bool inCoreOnly=false)
	Convert the group from a PointDataGrid. More...
template<typename PositionWrapper , typename InterrupterT = openvdb::util::NullInterrupter, typename VecT = typename internal::ValueTypeTraits<PositionWrapper>::Type>
float	computeVoxelSize (const PositionWrapper &positions, const uint32_t pointsPerVoxel, const math::Mat4d transform=math::Mat4d::identity(), const Index decimalPlaces=5, InterrupterT *const interrupter=nullptr)
template<typename PointDataTreeT , typename FilterT = NullFilter>
Index64	pointCount (const PointDataTreeT &tree, const FilterT &filter=NullFilter(), const bool inCoreOnly=false, const bool threaded=true)
	Count the total number of points in a PointDataTree. More...
template<typename PointDataTreeT , typename FilterT = NullFilter>
Index64	pointOffsets (std::vector< Index64 > &pointOffsets, const PointDataTreeT &tree, const FilterT &filter=NullFilter(), const bool inCoreOnly=false, const bool threaded=true)
	Populate an array of cumulative point offsets per leaf node. More...
template<typename PointDataGridT , typename GridT = typename PointDataGridT::template ValueConverter<Int32>::Type, typename FilterT = NullFilter>
GridT::Ptr	pointCountGrid (const PointDataGridT &grid, const FilterT &filter=NullFilter())
	Generate a new grid with voxel values to store the number of points per voxel. More...
template<typename PointDataGridT , typename GridT = typename PointDataGridT::template ValueConverter<Int32>::Type, typename FilterT = NullFilter>
GridT::Ptr	pointCountGrid (const PointDataGridT &grid, const openvdb::math::Transform &transform, const FilterT &filter=NullFilter())
	Generate a new grid that uses the supplied transform with voxel values to store the number of points per voxel. More...
template<typename PointDataTreeT >
AttributeSet::Descriptor::Ptr	makeDescriptorUnique (PointDataTreeT &tree)
	Deep copy the descriptor across all leaf nodes. More...
template<typename PointDataTreeT >
void	setStreamingMode (PointDataTreeT &tree, bool on=true)
	Toggle the streaming mode on all attributes in the tree to collapse the attributes after deconstructing a bound AttributeHandle to each array. This results in better memory efficiency when the data is streamed into another data structure (typically for rendering). More...
template<typename PointDataTreeT >
void	prefetch (PointDataTreeT &tree, bool position=true, bool otherAttributes=true)
	Sequentially pre-fetch all delayed-load voxel and attribute data from disk in order to accelerate subsequent random access. More...
template<typename PointDataTreeT >
void	deleteFromGroups (PointDataTreeT &pointTree, const std::vector< std::string > &groups, bool invert=false, bool drop=true)
	Delete points that are members of specific groups. More...
template<typename PointDataTreeT >
void	deleteFromGroup (PointDataTreeT &pointTree, const std::string &group, bool invert=false, bool drop=true)
	Delete points that are members of a group. More...
void	deleteMissingPointGroups (std::vector< std::string > &groups, const AttributeSet::Descriptor &descriptor)
	Delete any group that is not present in the Descriptor. More...
template<typename PointDataTreeT >
void	appendGroup (PointDataTreeT &tree, const Name &group)
	Appends a new empty group to the VDB tree. More...
template<typename PointDataTreeT >
void	appendGroups (PointDataTreeT &tree, const std::vector< Name > &groups)
	Appends new empty groups to the VDB tree. More...
template<typename PointDataTreeT >
void	dropGroup (PointDataTreeT &tree, const Name &group, const bool compact=true)
	Drops an existing group from the VDB tree. More...
template<typename PointDataTreeT >
void	dropGroups (PointDataTreeT &tree, const std::vector< Name > &groups)
	Drops existing groups from the VDB tree, the tree is compacted after dropping. More...
template<typename PointDataTreeT >
void	dropGroups (PointDataTreeT &tree)
	Drops all existing groups from the VDB tree, the tree is compacted after dropping. More...
template<typename PointDataTreeT >
void	compactGroups (PointDataTreeT &tree)
	Compacts existing groups of a VDB Tree to use less memory if possible. More...
template<typename PointDataTreeT , typename PointIndexTreeT >
void	setGroup (PointDataTreeT &tree, const PointIndexTreeT &indexTree, const std::vector< short > &membership, const Name &group, const bool remove=false)
	Sets group membership from a PointIndexTree-ordered vector. More...
template<typename PointDataTreeT >
void	setGroup (PointDataTreeT &tree, const Name &group, const bool member=true)
	Sets membership for the specified group for all points (on/off). More...
template<typename PointDataTreeT , typename FilterT >
void	setGroupByFilter (PointDataTreeT &tree, const Name &group, const FilterT &filter)
	Sets group membership based on a provided filter. More...
template<typename PointDataTreeT , typename MaskTreeT = typename PointDataTreeT::template ValueConverter<bool>::Type, typename FilterT = NullFilter>
std::enable_if< std::is_base_of< TreeBase, PointDataTreeT >::value &&std::is_same< typename MaskTreeT::ValueType, bool >::value, typename MaskTreeT::Ptr >::type	convertPointsToMask (const PointDataTreeT &tree, const FilterT &filter=NullFilter(), bool threaded=true)
	Extract a Mask Tree from a Point Data Tree. More...
template<typename PointDataGridT , typename MaskGridT = typename PointDataGridT::template ValueConverter<bool>::Type, typename FilterT = NullFilter>
std::enable_if< std::is_base_of< GridBase, PointDataGridT >::value &&std::is_same< typename MaskGridT::ValueType, bool >::value, typename MaskGridT::Ptr >::type	convertPointsToMask (const PointDataGridT &grid, const FilterT &filter=NullFilter(), bool threaded=true)
	Extract a Mask Grid from a Point Data Grid. More...
template<typename PointDataGridT , typename MaskT = typename PointDataGridT::template ValueConverter<bool>::Type, typename FilterT = NullFilter>
std::enable_if< std::is_same< typename MaskT::ValueType, bool >::value, typename MaskT::Ptr >::type	convertPointsToMask (const PointDataGridT &grid, const openvdb::math::Transform &transform, const FilterT &filter=NullFilter(), bool threaded=true)
	Extract a Mask Grid from a Point Data Grid using a new transform. More...
template<typename PointDataGridT , typename DeformerT , typename FilterT = NullFilter>
void	movePoints (PointDataGridT &points, DeformerT &deformer, const FilterT &filter=NullFilter(), future::Advect *objectNotInUse=nullptr, bool threaded=true)
	Move points in a PointDataGrid using a custom deformer. More...
template<typename PointDataGridT , typename DeformerT , typename FilterT = NullFilter>
void	movePoints (PointDataGridT &points, const math::Transform &transform, DeformerT &deformer, const FilterT &filter=NullFilter(), future::Advect *objectNotInUse=nullptr, bool threaded=true)
	Move points in a PointDataGrid using a custom deformer and a new transform. More...
template<typename PointDataGridT , typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
SdfT::Ptr	rasterizeSpheres (const PointDataGridT &points, const Real radius, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Narrow band sphere stamping with a uniform radius. More...
template<typename PointDataGridT , typename RadiusT = float, typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
SdfT::Ptr	rasterizeSpheres (const PointDataGridT &points, const std::string &radius, const Real scale=1.0, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Narrow band sphere stamping with a varying radius. More...
template<typename PointDataGridT , typename AttributeTypes , typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
GridPtrVec	rasterizeSpheres (const PointDataGridT &points, const Real radius, const std::vector< std::string > &attributes, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Narrow band sphere stamping with a uniform radius and closest point attribute transfer. More...
template<typename PointDataGridT , typename AttributeTypes , typename RadiusT = float, typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
GridPtrVec	rasterizeSpheres (const PointDataGridT &points, const std::string &radius, const std::vector< std::string > &attributes, const Real scale=1.0, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Narrow band sphere stamping with a varying radius and closest point attribute transfer. More...
template<typename PointDataGridT , typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
SdfT::Ptr	rasterizeSmoothSpheres (const PointDataGridT &points, const Real radius, const Real searchRadius, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Smoothed point distribution based sphere stamping with a uniform radius. More...
template<typename PointDataGridT , typename RadiusT = float, typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
SdfT::Ptr	rasterizeSmoothSpheres (const PointDataGridT &points, const std::string &radius, const Real radiusScale, const Real searchRadius, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Smoothed point distribution based sphere stamping with a varying radius. More...
template<typename PointDataGridT , typename AttributeTypes , typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
GridPtrVec	rasterizeSmoothSpheres (const PointDataGridT &points, const Real radius, const Real searchRadius, const std::vector< std::string > &attributes, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Smoothed point distribution based sphere stamping with a uniform radius and closest point attribute transfer. More...
template<typename PointDataGridT , typename AttributeTypes , typename RadiusT = float, typename SdfT = typename PointDataGridT::template ValueConverter<float>::Type, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
GridPtrVec	rasterizeSmoothSpheres (const PointDataGridT &points, const std::string &radius, const Real radiusScale, const Real searchRadius, const std::vector< std::string > &attributes, const Real halfband=LEVEL_SET_HALF_WIDTH, math::Transform::Ptr transform=nullptr, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Smoothed point distribution based sphere stamping with a varying radius and closest point attribute transfer. More...
template<bool Staggered, typename ValueT , typename FilterT = NullFilter, typename PointDataTreeT = PointDataTree>
auto	rasterizeTrilinear (const PointDataTreeT &points, const std::string &attribute, const FilterT &filter=NullFilter())
	Perform weighted trilinear rasterization of all points within a voxel. This method takes and returns a tree i.e. ignores grid transformations. More...
template<typename PointDataGridT >
PointDataGridT::Ptr	replicate (const PointDataGridT &source, const Index multiplier, const std::vector< std::string > &attributes, const std::string &scaleAttribute="", const std::string &replicationIndex="")
	Replicates points provided in a source grid into a new grid, transfering and creating attributes found in a provided attribute vector. If an attribute doesn't exist, it is ignored. Position is always replicated, leaving the new points exactly over the top of the source points. More...
template<typename PointDataGridT >
PointDataGridT::Ptr	replicate (const PointDataGridT &source, const Index multiplier, const std::string &scaleAttribute="", const std::string &replicationIndex="")
	Replicates points provided in a source grid into a new grid, transfering and creating all attributes from the source grid. Position is always replicated, leaving the new points exactly over the top of the source points. More...
template<typename PointDataGridT , typename SourceGridT , typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
void	pointSample (PointDataGridT &points, const SourceGridT &sourceGrid, const Name &targetAttribute="", const FilterT &filter=NullFilter(), InterrupterT *const interrupter=nullptr)
	Performs closest point sampling from a VDB grid onto a VDB Points attribute. More...
template<typename PointDataGridT , typename SourceGridT , typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
void	boxSample (PointDataGridT &points, const SourceGridT &sourceGrid, const Name &targetAttribute="", const FilterT &filter=NullFilter(), InterrupterT *const interrupter=nullptr)
	Performs tri-linear sampling from a VDB grid onto a VDB Points attribute. More...
template<typename PointDataGridT , typename SourceGridT , typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
void	quadraticSample (PointDataGridT &points, const SourceGridT &sourceGrid, const Name &targetAttribute="", const FilterT &filter=NullFilter(), InterrupterT *const interrupter=nullptr)
	Performs tri-quadratic sampling from a VDB grid onto a VDB Points attribute. More...
template<typename PointDataGridT , typename SourceGridT , typename TargetValueT = DummySampleType, typename SamplerT = SampleWithRounding, typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
void	sampleGrid (size_t order, PointDataGridT &points, const SourceGridT &sourceGrid, const Name &targetAttribute, const FilterT &filter=NullFilter(), const SamplerT &sampler=SampleWithRounding(), InterrupterT *const interrupter=nullptr, const bool threaded=true)
	Performs sampling and conversion from a VDB grid onto a VDB Points attribute. More...
template<typename GridT , typename RandGenT = std::mt19937, typename PositionArrayT = TypedAttributeArray<Vec3f, NullCodec>, typename PointDataGridT = Grid< typename points::TreeConverter<typename GridT::TreeType>::Type>, typename InterrupterT = util::NullInterrupter>
PointDataGridT::Ptr	uniformPointScatter (const GridT &grid, const Index64 count, const unsigned int seed=0, const float spread=1.0f, InterrupterT *interrupter=nullptr)
	The free functions depend on the following class: More...
template<typename GridT , typename RandGenT = std::mt19937, typename PositionArrayT = TypedAttributeArray<Vec3f, NullCodec>, typename PointDataGridT = Grid< typename points::TreeConverter<typename GridT::TreeType>::Type>, typename InterrupterT = util::NullInterrupter>
PointDataGridT::Ptr	denseUniformPointScatter (const GridT &grid, const float pointsPerVoxel, const unsigned int seed=0, const float spread=1.0f, InterrupterT *interrupter=nullptr)
	Uniformly scatter a fixed number of points per active voxel. If the pointsPerVoxel value provided is a fractional value, each voxel calculates a delta value of how likely it is to contain an extra point. More...
template<typename GridT , typename RandGenT = std::mt19937, typename PositionArrayT = TypedAttributeArray<Vec3f, NullCodec>, typename PointDataGridT = Grid< typename points::TreeConverter<typename GridT::TreeType>::Type>, typename InterrupterT = util::NullInterrupter>
PointDataGridT::Ptr	nonUniformPointScatter (const GridT &grid, const float pointsPerVoxel, const unsigned int seed=0, const float spread=1.0f, InterrupterT *interrupter=nullptr)
	Non uniformly scatter points per active voxel. The pointsPerVoxel value is used to weight each grids cell value to compute a fixed number of points for every active voxel. If the computed result is a fractional value, each voxel calculates a delta value of how likely it is to contain an extra point. More...
template<typename ValueT , typename CodecT = UnknownCodec, typename FilterT = NullFilter, typename PointDataTreeT >
std::pair< ValueT, ValueT >	evalMinMax (const PointDataTreeT &points, const std::string &attribute, const FilterT &filter=NullFilter())
	Evaluates the minimum and maximum values of a point attribute. More...
template<typename ValueT , typename CodecT = UnknownCodec, typename FilterT = NullFilter, typename PointDataTreeT >
ConvertElementType< ValueT, double >::Type	evalAverage (const PointDataTreeT &points, const std::string &attribute, const FilterT &filter=NullFilter())
	Evaluates the average value of a point attribute. More...
template<typename ValueT , typename CodecT = UnknownCodec, typename FilterT = NullFilter, typename PointDataTreeT >
PromoteType< ValueT >::Highest	accumulate (const PointDataTreeT &points, const std::string &attribute, const FilterT &filter=NullFilter())
	Evaluates the total value of a point attribute. More...
template<typename ValueT , typename CodecT = UnknownCodec, typename FilterT = NullFilter, typename PointDataTreeT >
bool	evalMinMax (const PointDataTreeT &points, const std::string &attribute, ValueT &min, ValueT &max, const FilterT &filter=NullFilter(), typename PointDataTreeT::template ValueConverter< ValueT >::Type *minTree=nullptr, typename PointDataTreeT::template ValueConverter< ValueT >::Type *maxTree=nullptr)
	Evaluates the minimum and maximum values of a point attribute and returns whether the values are valid. Optionally constructs localised min and max value trees. More...
template<typename ValueT , typename CodecT = UnknownCodec, typename FilterT = NullFilter, typename PointDataTreeT , typename ResultTreeT = typename ConvertElementType<ValueT, double>::Type>
bool	evalAverage (const PointDataTreeT &points, const std::string &attribute, typename ConvertElementType< ValueT, double >::Type &average, const FilterT &filter=NullFilter(), typename PointDataTreeT::template ValueConverter< ResultTreeT >::Type *averageTree=nullptr)
	Evaluates the average value of a point attribute and returns whether the value is valid. Optionally constructs localised average value trees. More...
template<typename ValueT , typename CodecT = UnknownCodec, typename FilterT = NullFilter, typename PointDataTreeT , typename ResultTreeT = typename PromoteType<ValueT>::Highest>
bool	accumulate (const PointDataTreeT &points, const std::string &attribute, typename PromoteType< ValueT >::Highest &total, const FilterT &filter=NullFilter(), typename PointDataTreeT::template ValueConverter< ResultTreeT >::Type *totalTree=nullptr)
	Evaluates the total value of a point attribute and returns whether the value is valid. Optionally constructs localised total value trees. More...
template<typename PointDataTreeOrGridT , typename TransferT , typename FilterT = NullFilter, typename InterrupterT = util::NullInterrupter>
void	rasterize (const PointDataTreeOrGridT &points, TransferT &transfer, const FilterT &filter=NullFilter(), InterrupterT *interrupter=nullptr)
	Perform potentially complex rasterization from a user defined transfer scheme. More...
template<typename PointDataTreeT >
void	setGroupByRandomTarget (PointDataTreeT &tree, const Name &group, const Index64 targetPoints, const unsigned int seed=0)
template<typename PointDataTreeT >
void	setGroupByRandomPercentage (PointDataTreeT &tree, const Name &group, const float percentage=10.0f, const unsigned int seed=0)

Typedef Documentation

using ActiveFilter = ValueMaskFilter<true>

using GroupAttributeArray = TypedAttributeArray<GroupType, GroupCodec>

using GroupType = uint8_t

using InactiveFilter = ValueMaskFilter<false>

using PointDataGrid = Grid<PointDataTree>

Point data grid.

using PointDataTree = tree::Tree<tree::RootNode<tree::InternalNode<tree::InternalNode <PointDataLeafNode<PointDataIndex32, 3>, 4>, 5>>>

Point index tree configured to match the default VDB configurations.

using StringAttributeArray = TypedAttributeArray<Index, StringCodec<false>>

Enumeration Type Documentation

enum RasterMode

strong

How to composite points into a volume.

Enumerator
ACCUMULATE
MAXIMUM
AVERAGE

Function Documentation

PromoteType< ValueT >::Highest accumulate	(	const PointDataTreeT &	points,
		const std::string &	attribute,
		const FilterT &	filter = NullFilter()
	)

Evaluates the total value of a point attribute.

Performs parallel reduction by summing all values. The reduction arithmetic and return value precision evaluates to: PromoteType<ValueT>::Highest which, for POD and VDB math types, is ValueT at its highest bit precision. If the PointDataGrid is empty or the filter evalutes to empty, zeroVal<ValueT>() is returned.

Note

The PromoteType of the attribute must be copy constructible, support the same type + operator. This method will throw if ValueT does not match the given attribute. The function is deterministic.

Warning

if "P" is provided, the result is undefined.

Parameters

points	the point tree
attribute	the attribute to reduce
filter	a filter to apply to points

Returns

the total value

bool accumulate	(	const PointDataTreeT &	points,
		const std::string &	attribute,
		typename PromoteType< ValueT >::Highest &	total,
		const FilterT &	filter = NullFilter(),
		typename PointDataTreeT::template ValueConverter< ResultTreeT >::Type *	totalTree = nullptr
	)

Evaluates the total value of a point attribute and returns whether the value is valid. Optionally constructs localised total value trees.

Performs parallel reduction by summing all values. The reduction arithmetic and return value precision evaluates to: PromoteType<ValueT>::Highest which, for POD and VDB math types, is ValueT at its highest bit precision. This method will return true total has been set, false otherwise (when no points existed or a filter evaluated to empty).

Note

The PromoteType of the attribute must be copy constructible, support the same type + operator. This method will throw if ValueT does not match the given attribute. The function is deterministic.

Warning

if "P" is provided, the result is undefined.

Parameters

points	the point tree
attribute	the attribute to reduce
total	the computed total value
filter	a filter to apply to points
totalTree	if provided, builds a tiled tree of localised total results.

Returns

true if total has been set, false otherwise. Can be false if no points were processed or if the tree was empty.

Example:

using namespace openvdb;
using namespace openvdb::points;

// accumulate and store per leaf values in a new tree
PromoteType<uint8_t>::Highest total;  // evaluates to uint64_t
PointDataTree::ValueConverter<decltype(total)>::Type totalTree; // uint64_t tree of totals
bool success = accumulate<uint8_t>(tree, "attrib", total, NullFilter(), &totalTree);

void advectPoints ( PointDataGridT &  points, const VelGridT &  velocity, const Index  integrationOrder, const double  dt, const Index  timeSteps, const AdvectFilterT &  advectFilter = NullFilter(), const FilterT &  filter = NullFilter(), const bool  cached = true  )

inline

Advect points in a PointDataGrid through a velocity grid.

Parameters

points	the PointDataGrid containing the points to be advected.
velocity	a velocity grid to be sampled.
integrationOrder	the integration scheme to use (1 is forward euler, 4 is runge-kutta 4th)
dt	delta time.
timeSteps	number of advection steps to perform.
advectFilter	an optional advection index filter (moves a subset of the points)
filter	an optional index filter (deletes a subset of the points)
cached	caches velocity interpolation for faster performance, disable to use less memory (default is on).

void appendAttribute ( PointDataTreeT &  tree, const Name &  name, const NamePair &  type, const Index  strideOrTotalSize = 1, const bool  constantStride = true, const Metadata *  defaultValue = nullptr, const bool  hidden = false, const bool  transient = false  )

inline

Appends a new attribute to the VDB tree (this method does not require a templated AttributeType)

Parameters

tree	the PointDataTree to be appended to.
name	name for the new attribute.
type	the type of the attibute.
strideOrTotalSize	the stride of the attribute
constantStride	if false, stride is interpreted as total size of the array
defaultValue	metadata default attribute value
hidden	mark attribute as hidden
transient	mark attribute as transient

void appendAttribute ( PointDataTreeT &  tree, const std::string &  name, const ValueType &  uniformValue = point_attribute_internal::Default<ValueType>::value(), const Index  strideOrTotalSize = 1, const bool  constantStride = true, const TypedMetadata< ValueType > *  defaultValue = nullptr, const bool  hidden = false, const bool  transient = false  )

inline

Appends a new attribute to the VDB tree.

Parameters

tree	the PointDataTree to be appended to.
name	name for the new attribute
uniformValue	the initial value of the attribute
strideOrTotalSize	the stride of the attribute
constantStride	if false, stride is interpreted as total size of the array
defaultValue	metadata default attribute value
hidden	mark attribute as hidden
transient	mark attribute as transient

void appendGroup ( PointDataTreeT &  tree, const Name &  group  )

inline

Appends a new empty group to the VDB tree.

Parameters

tree	the PointDataTree to be appended to.
group	name of the new group.

void appendGroups ( PointDataTreeT &  tree, const std::vector< Name > &  groups  )

inline

Appends new empty groups to the VDB tree.

Parameters

tree	the PointDataTree to be appended to.
groups	names of the new groups.

void boxSample ( PointDataGridT &  points, const SourceGridT &  sourceGrid, const Name &  targetAttribute = "", const FilterT &  filter = NullFilter(), InterrupterT *const  interrupter = nullptr  )

inline

Performs tri-linear sampling from a VDB grid onto a VDB Points attribute.

Parameters

points	the PointDataGrid whose points will be sampled on to
sourceGrid	VDB grid which will be sampled
targetAttribute	a target attribute on the points which will hold samples. This attribute will be created with the source grid type if it does not exist, and with the source grid name if the name is empty
filter	an optional index filter
interrupter	an optional interrupter

Note

The target attribute may exist provided it can be cast to the SourceGridT ValueType

void collapseAttribute ( PointDataTreeT &  tree, const Name &  name, const ValueType &  uniformValue = point_attribute_internal::Default<ValueType>::value()  )

inline

Collapse the attribute into a uniform value.

Parameters

tree	the PointDataTree in which to collapse the attribute.
name	name for the attribute.
uniformValue	value of the attribute

void compactAttributes ( PointDataTreeT &  tree )

inline

Compact attributes in a VDB tree (if possible).

Parameters

tree	the PointDataTree.

void compactGroups ( PointDataTreeT &  tree )

inline

Compacts existing groups of a VDB Tree to use less memory if possible.

Parameters

tree	the PointDataTree to be compacted.

float computeVoxelSize ( const PositionWrapper &  positions, const uint32_t  pointsPerVoxel, const math::Mat4d  transform = math::Mat4d::identity(), const Index  decimalPlaces = 5, InterrupterT *const  interrupter = nullptr  )

inline

@ brief Given a container of world space positions and a target points per voxel, compute a uniform voxel size that would best represent the storage of the points in a grid. This voxel size is typically used for conversion of the points into a PointDataGrid.

Parameters

positions	array of world space positions
pointsPerVoxel	the target number of points per voxel, must be positive and non-zero
transform	voxel size will be computed using this optional transform if provided
decimalPlaces	for readability, truncate voxel size to this number of decimals
interrupter	an optional interrupter

Note

VecT will be PositionWrapper::value_type or Vec3R (if there is no value_type defined)

if none or one point provided in positions, the default voxel size of 0.1 will be returned

void convertPointDataGridAttribute ( TypedAttribute &  attribute, const PointDataTreeT &  tree, const std::vector< Index64 > &  pointOffsets, const Index64  startOffset, const unsigned  arrayIndex, const Index  stride = 1, const FilterT &  filter = NullFilter(), const bool  inCoreOnly = false  )

inline

Convert the attribute from a PointDataGrid.

Parameters

attribute	the attribute to be populated.
tree	the PointDataTree to be converted.
pointOffsets	a vector of cumulative point offsets for each leaf.
startOffset	a value to shift all the point offsets by
arrayIndex	the index in the Descriptor of the array to be converted.
stride	the stride of the attribute
filter	an index filter
inCoreOnly	true if out-of-core leaf nodes are to be ignored

void convertPointDataGridGroup ( Group &  group, const PointDataTreeT &  tree, const std::vector< Index64 > &  pointOffsets, const Index64  startOffset, const AttributeSet::Descriptor::GroupIndex  index, const FilterT &  filter = NullFilter(), const bool  inCoreOnly = false  )

inline

Convert the group from a PointDataGrid.

Parameters

group	the group to be populated.
tree	the PointDataTree to be converted.
pointOffsets	a vector of cumulative point offsets for each leaf
startOffset	a value to shift all the point offsets by
index	the group index to be converted.
filter	an index filter
inCoreOnly	true if out-of-core leaf nodes are to be ignored

void convertPointDataGridPosition ( PositionAttribute &  positionAttribute, const PointDataGridT &  grid, const std::vector< Index64 > &  pointOffsets, const Index64  startOffset, const FilterT &  filter = NullFilter(), const bool  inCoreOnly = false  )

inline

Convert the position attribute from a Point Data Grid.

Parameters

positionAttribute	the position attribute to be populated.
grid	the PointDataGrid to be converted.
pointOffsets	a vector of cumulative point offsets for each leaf
startOffset	a value to shift all the point offsets by
filter	an index filter
inCoreOnly	true if out-of-core leaf nodes are to be ignored

std::enable_if< std::is_base_of< TreeBase, PointDataTreeT >::value &&std::is_same< typename MaskTreeT::ValueType, bool >::value, typename MaskTreeT::Ptr >::type convertPointsToMask ( const PointDataTreeT &  tree, const FilterT &  filter = NullFilter(), bool  threaded = true  )

inline

Extract a Mask Tree from a Point Data Tree.

Parameters

tree	the PointDataTree to extract the mask from.
filter	an optional index filter
threaded	enable or disable threading (threading is enabled by default)

std::enable_if< std::is_base_of< GridBase, PointDataGridT >::value &&std::is_same< typename MaskGridT::ValueType, bool >::value, typename MaskGridT::Ptr >::type convertPointsToMask ( const PointDataGridT &  grid, const FilterT &  filter = NullFilter(), bool  threaded = true  )

inline

Extract a Mask Grid from a Point Data Grid.

Parameters

grid	the PointDataGrid to extract the mask from.
filter	an optional index filter
threaded	enable or disable threading (threading is enabled by default)

Note

this method is only available for Bool Grids and Mask Grids

std::enable_if< std::is_same< typename MaskT::ValueType, bool >::value, typename MaskT::Ptr >::type convertPointsToMask ( const PointDataGridT &  grid, const openvdb::math::Transform &  transform, const FilterT &  filter = NullFilter(), bool  threaded = true  )

inline

Extract a Mask Grid from a Point Data Grid using a new transform.

Parameters

grid	the PointDataGrid to extract the mask from.
transform	target transform for the mask.
filter	an optional index filter
threaded	enable or disable threading (threading is enabled by default)

Note

this method is only available for Bool Grids and Mask Grids

PointDataGridT::Ptr createPointDataGrid ( const PointIndexGridT &  pointIndexGrid, const PositionArrayT &  positions, const math::Transform &  xform, const Metadata *  positionDefaultValue = nullptr  )

inline

Localises points with position into a PointDataGrid into two stages: allocation of the leaf attribute data and population of the positions.

Parameters

pointIndexGrid	a PointIndexGrid into the points.
positions	list of world space point positions.
xform	world to index space transform.
positionDefaultValue	metadata default position value

Note

The position data must be supplied in a Point-Partitioner compatible data structure. A convenience PointAttributeVector class is offered.

The position data is populated separately to perform world space to voxel space conversion and apply quantisation.

A PointIndexGrid to the points must be supplied to perform this operation. Typically this is built implicitly by the PointDataGrid constructor.

PointDataGridT::Ptr createPointDataGrid ( const std::vector< ValueT > &  positions, const math::Transform &  xform, const Metadata *  positionDefaultValue = nullptr  )

inline

Convenience method to create a PointDataGrid from a std::vector of point positions.

Parameters

positions	list of world space point positions.
xform	world to index space transform.
positionDefaultValue	metadata default position value

Note

This method implicitly wraps the std::vector for a Point-Partitioner compatible data structure and creates the required PointIndexGrid to the points.

void deleteFromGroup ( PointDataTreeT &  pointTree, const std::string &  group, bool  invert = false, bool  drop = true  )

inline

Delete points that are members of a group.

This method will delete points which are members of the supplied group and will optionally drop the group from the tree. An invert flag can be used to delete points that belong to none of the groups.

Parameters

pointTree	the point tree with the group to delete
group	the name of the group to delete
invert	if enabled, points not belonging to any of the groups will be deleted
drop	if enabled and invert is disabled, the group will be dropped from the tree

Note

If the invert flag is true, the group will not be dropped after deleting points regardless of the value of the drop parameter.

void deleteFromGroups ( PointDataTreeT &  pointTree, const std::vector< std::string > &  groups, bool  invert = false, bool  drop = true  )

inline

Delete points that are members of specific groups.

This method will delete points which are members of any of the supplied groups and will optionally drop the groups from the tree. An invert flag can be used to delete points that belong to none of the groups.

Parameters

pointTree	the point tree
groups	the groups from which to delete points
invert	if enabled, points not belonging to any of the groups will be deleted
drop	if enabled and invert is disabled, the groups will be dropped from the tree

Note

If the invert flag is true, none of the groups will be dropped after deleting points regardless of the value of the drop parameter.

void deleteMissingPointGroups ( std::vector< std::string > &  groups, const AttributeSet::Descriptor &  descriptor  )

inline

Delete any group that is not present in the Descriptor.

Parameters

groups	the vector of group names.
descriptor	the descriptor that holds the group map.

PointDataGridT::Ptr denseUniformPointScatter ( const GridT &  grid, const float  pointsPerVoxel, const unsigned int  seed = 0, const float  spread = 1.0f, InterrupterT *  interrupter = nullptr  )

inline

Uniformly scatter a fixed number of points per active voxel. If the pointsPerVoxel value provided is a fractional value, each voxel calculates a delta value of how likely it is to contain an extra point.

Parameters

grid	A source grid. The resulting PointDataGrid will copy this grids transform and scatter in its active voxelized topology.
pointsPerVoxel	The number of points to scatter per voxel
seed	A seed for the RandGenT
spread	The spread of points as a scale from each voxels center. A value of 1.0f indicates points can be placed anywhere within the voxel, where as a value of 0.0f will force all points to be created exactly at the centers of each voxel.
interrupter	An optional interrupter

Note

returns the scattered PointDataGrid

void dropAttribute ( PointDataTreeT &  tree, const size_t &  index  )

inline

Drop one attribute from the VDB tree (convenience method).

Parameters

tree	the PointDataTree to be dropped from.
index	index of the attribute to drop.

void dropAttribute ( PointDataTreeT &  tree, const Name &  name  )

inline

Drop one attribute from the VDB tree (convenience method).

Parameters

tree	the PointDataTree to be dropped from.
name	name of the attribute to drop.

void dropAttributes ( PointDataTreeT &  tree, const std::vector< size_t > &  indices  )

inline

Drops attributes from the VDB tree.

Parameters

tree	the PointDataTree to be dropped from.
indices	indices of the attributes to drop.

void dropAttributes ( PointDataTreeT &  tree, const std::vector< Name > &  names  )

inline

Drops attributes from the VDB tree.

Parameters

tree	the PointDataTree to be dropped from.
names	names of the attributes to drop.

void dropGroup ( PointDataTreeT &  tree, const Name &  group, const bool  compact = true  )

inline

Drops an existing group from the VDB tree.

Parameters

tree	the PointDataTree to be dropped from.
group	name of the group.
compact	compact attributes if possible to reduce memory - if dropping more than one group, compacting once at the end will be faster

void dropGroups ( PointDataTreeT &  tree, const std::vector< Name > &  groups  )

inline

Drops existing groups from the VDB tree, the tree is compacted after dropping.

Parameters

tree	the PointDataTree to be dropped from.
groups	names of the groups.

void dropGroups ( PointDataTreeT &  tree )

inline

Drops all existing groups from the VDB tree, the tree is compacted after dropping.

Parameters

tree	the PointDataTree to be dropped from.

ConvertElementType< ValueT, double >::Type evalAverage	(	const PointDataTreeT &	points,
		const std::string &	attribute,
		const FilterT &	filter = NullFilter()
	)

Evaluates the average value of a point attribute.

Performs parallel reduction by cumulative moving average. The reduction arithmetic and return value precision evaluates to: ConvertElementType<ValueT, double>::Type which, for POD and VDB math types, is ValueT at double precision. If the PointDataGrid is empty or the filter evalutes to empty, zeroVal<ValueT>() is returned.

Note

The ConvertElementType of the attribute must be copy constructible, support the same type + - * operators and * / operators from a double. This method will throw if ValueT does not match the given attribute. The function is deterministic.

Warning

if "P" is provided, the result is undefined.

Parameters

points	the point tree
attribute	the attribute to reduce
filter	a filter to apply to points

Returns

the average value

bool evalAverage	(	const PointDataTreeT &	points,
		const std::string &	attribute,
		typename ConvertElementType< ValueT, double >::Type &	average,
		const FilterT &	filter = NullFilter(),
		typename PointDataTreeT::template ValueConverter< ResultTreeT >::Type *	averageTree = nullptr
	)

Evaluates the average value of a point attribute and returns whether the value is valid. Optionally constructs localised average value trees.

Performs parallel reduction by cumulative moving average. The reduction arithmetic and return value precision evaluates to: ConvertElementType<ValueT, double>::Type which, for POD and VDB math types, is ValueT at double precision. This method will return true average has been set, false otherwise (when no points existed or a filter evaluated to empty).

Note

The ConvertElementType of the attribute must be copy constructible, support the same type + - * operators and * / operators from a double. This method will throw if ValueT does not match the given attribute. The function is deterministic.

Warning

if "P" is provided, the result is undefined.

Parameters

points	the point tree
attribute	the attribute to reduce
average	the computed averaged value at double precision
filter	a filter to apply to points
averageTree	if provided, builds a tiled tree of localised avg results.

Returns

true if average has been set, false otherwise. Can be false if no points were processed or if the tree was empty.

Example:

using namespace openvdb;
using namespace openvdb::points

// average and store per leaf values in a new tree
ConvertElementType<uint8_t, double>::Type avg;  // evaluates to double
PointDataTree::ValueConverter<decltype(avg)>::Type avgTree; // double tree of averages
bool success = evalAverage<uint8_t>(tree, "attrib", avg, NullFilter(), &avgTree);

std::pair< ValueT, ValueT > evalMinMax	(	const PointDataTreeT &	points,
		const std::string &	attribute,
		const FilterT &	filter = NullFilter()
	)

Evaluates the minimum and maximum values of a point attribute.

Performs parallel reduction by comparing values using their less than and greater than operators. If the PointDataGrid is empty or the filter evalutes to empty, zeroVal<ValueT>() is returned for both values.

Note

The ValueT of the attribute must be copy constructible. This method will throw if the templated ValueT does not match the given attribute. For vectors and matrices, this results in per component comparisons. See evalExtents for magnitudes or more custom control.

Warning

if "P" is provided, the result is undefined.

Parameters

points	the point tree
attribute	the attribute to reduce
filter	a filter to apply to points

Returns

min,max value pair

bool evalMinMax	(	const PointDataTreeT &	points,
		const std::string &	attribute,
		ValueT &	min,
		ValueT &	max,
		const FilterT &	filter = NullFilter(),
		typename PointDataTreeT::template ValueConverter< ValueT >::Type *	minTree = nullptr,
		typename PointDataTreeT::template ValueConverter< ValueT >::Type *	maxTree = nullptr
	)

Evaluates the minimum and maximum values of a point attribute and returns whether the values are valid. Optionally constructs localised min and max value trees.

Performs parallel reduction by comparing values using their less than and greater than operators. This method will return true if min and max have been set, false otherwise (when no points existed or a filter evaluated to empty).

Note

The ValueT of the attribute must also be copy constructible. This method will throw if the templated ValueT does not match the given attribute. For vectors and matrices, this results in per component comparisons. See evalExtents for magnitudes or more custom control.

Warning

if "P" is provided, the result is undefined.

Parameters

points	the point tree
attribute	the attribute to reduce
min	the computed min value
max	the computed max value
filter	a filter to apply to points
minTree	if provided, builds a tiled tree of localised min results
maxTree	if provided, builds a tiled tree of localised max results

Returns

true if min and max have been set, false otherwise. Can be false if no points were processed or if the tree was empty.

FloatT openvdb::v12_0::points::fixedPointToFloatingPoint ( const IntegerT  s )

inline

FloatVectorT openvdb::v12_0::points::fixedPointToFloatingPoint ( const math::Vec3< IntegerT > &  v )

inline

IntegerT openvdb::v12_0::points::floatingPointToFixedPoint ( const FloatT  s )

inline

IntegerVectorT openvdb::v12_0::points::floatingPointToFixedPoint ( const math::Vec3< FloatT > &  v )

inline

bool openvdb::v12_0::points::isGroup ( const AttributeArray &  array )

inline

bool openvdb::v12_0::points::isString ( const AttributeArray &  array )

inline

Index64 iterCount ( const IterT &  iter )

inline

Count up the number of times the iterator can iterate.

Parameters

iter	the iterator.

Note

counting by iteration only performed where a dynamic filter is in use,

AttributeSet::Descriptor::Ptr makeDescriptorUnique ( PointDataTreeT &  tree )

inline

Deep copy the descriptor across all leaf nodes.

Parameters

tree	the PointDataTree.

Returns

the new descriptor.

Note

This method will fail if the Descriptors in the tree are not all identical.

void movePoints ( PointDataGridT &  points, DeformerT &  deformer, const FilterT &  filter = NullFilter(), future::Advect *  objectNotInUse = nullptr, bool  threaded = true  )

inline

Move points in a PointDataGrid using a custom deformer.

Parameters

points	the PointDataGrid containing the points to be moved.
deformer	a custom deformer that defines how to move the points.
filter	an optional index filter
objectNotInUse	for future use, this object is currently ignored
threaded	enable or disable threading (threading is enabled by default)

void movePoints ( PointDataGridT &  points, const math::Transform &  transform, DeformerT &  deformer, const FilterT &  filter = NullFilter(), future::Advect *  objectNotInUse = nullptr, bool  threaded = true  )

inline

Move points in a PointDataGrid using a custom deformer and a new transform.

Parameters

points	the PointDataGrid containing the points to be moved.
transform	target transform to use for the resulting points.
deformer	a custom deformer that defines how to move the points.
filter	an optional index filter
objectNotInUse	for future use, this object is currently ignored
threaded	enable or disable threading (threading is enabled by default)

PointDataGridT::Ptr nonUniformPointScatter ( const GridT &  grid, const float  pointsPerVoxel, const unsigned int  seed = 0, const float  spread = 1.0f, InterrupterT *  interrupter = nullptr  )

inline

Non uniformly scatter points per active voxel. The pointsPerVoxel value is used to weight each grids cell value to compute a fixed number of points for every active voxel. If the computed result is a fractional value, each voxel calculates a delta value of how likely it is to contain an extra point.

Parameters

grid	A source grid. The resulting PointDataGrid will copy this grids transform, voxelized topology and use its values to compute a target points per voxel. The grids ValueType must be convertible to a scalar value. Only active and larger than zero values will contain points.
pointsPerVoxel	The number of points to scatter per voxel
seed	A seed for the RandGenT
spread	The spread of points as a scale from each voxels center. A value of 1.0f indicates points can be placed anywhere within the voxel, where as a value of 0.0f will force all points to be created exactly at the centers of each voxel.
interrupter	An optional interrupter

Note

returns the scattered PointDataGrid

Index64 pointCount ( const PointDataTreeT &  tree, const FilterT &  filter = NullFilter(), const bool  inCoreOnly = false, const bool  threaded = true  )

inline

Count the total number of points in a PointDataTree.

Parameters

tree	the PointDataTree in which to count the points
filter	an optional index filter
inCoreOnly	if true, points in out-of-core leaf nodes are not counted
threaded	enable or disable threading (threading is enabled by default)

GridT::Ptr pointCountGrid ( const PointDataGridT &  grid, const FilterT &  filter = NullFilter()  )

inline

Generate a new grid with voxel values to store the number of points per voxel.

Parameters

grid	the PointDataGrid to use to compute the count grid
filter	an optional index filter

Note

The return type of the grid must be an integer or floating-point scalar grid.

GridT::Ptr pointCountGrid ( const PointDataGridT &  grid, const openvdb::math::Transform &  transform, const FilterT &  filter = NullFilter()  )

inline

Generate a new grid that uses the supplied transform with voxel values to store the number of points per voxel.

Parameters

grid	the PointDataGrid to use to compute the count grid
transform	the transform to use to compute the count grid
filter	an optional index filter

Note

The return type of the grid must be an integer or floating-point scalar grid.

Index64 pointOffsets ( std::vector< Index64 > &  pointOffsets, const PointDataTreeT &  tree, const FilterT &  filter = NullFilter(), const bool  inCoreOnly = false, const bool  threaded = true  )

inline

Populate an array of cumulative point offsets per leaf node.

Parameters

pointOffsets	array of offsets to be populated
tree	the PointDataTree from which to populate the offsets
filter	an optional index filter
inCoreOnly	if true, points in out-of-core leaf nodes are ignored
threaded	enable or disable threading (threading is enabled by default)

Returns

The final cumulative point offset.

void pointSample ( PointDataGridT &  points, const SourceGridT &  sourceGrid, const Name &  targetAttribute = "", const FilterT &  filter = NullFilter(), InterrupterT *const  interrupter = nullptr  )

inline

Performs closest point sampling from a VDB grid onto a VDB Points attribute.

Parameters

points	the PointDataGrid whose points will be sampled on to
sourceGrid	VDB grid which will be sampled
targetAttribute	a target attribute on the points which will hold samples. This attribute will be created with the source grid type if it does not exist, and with the source grid name if the name is empty
filter	an optional index filter
interrupter	an optional interrupter

Note

The target attribute may exist provided it can be cast to the SourceGridT ValueType

void populateAttribute ( PointDataTreeT &  tree, const PointIndexTreeT &  pointIndexTree, const openvdb::Name &  attributeName, const PointArrayT &  data, const Index  stride = 1, const bool  insertMetadata = true  )

inline

Stores point attribute data in an existing PointDataGrid attribute.

Parameters

tree	the PointDataGrid to be populated.
pointIndexTree	a PointIndexTree into the points.
attributeName	the name of the VDB Points attribute to be populated.
data	a wrapper to the attribute data.
stride	the stride of the attribute
insertMetadata	true if strings are to be automatically inserted as metadata.

Note

A PointIndexGrid to the points must be supplied to perform this operation. This is required to ensure the same point index ordering.

void prefetch ( PointDataTreeT &  tree, bool  position = true, bool  otherAttributes = true  )

inline

Sequentially pre-fetch all delayed-load voxel and attribute data from disk in order to accelerate subsequent random access.

Parameters

tree	the PointDataTree.
position	if enabled, prefetch the position attribute (default is on)
otherAttributes	if enabled, prefetch all other attributes (default is on)

void quadraticSample ( PointDataGridT &  points, const SourceGridT &  sourceGrid, const Name &  targetAttribute = "", const FilterT &  filter = NullFilter(), InterrupterT *const  interrupter = nullptr  )

inline

Performs tri-quadratic sampling from a VDB grid onto a VDB Points attribute.

Parameters

points	the PointDataGrid whose points will be sampled on to
sourceGrid	VDB grid which will be sampled
targetAttribute	a target attribute on the points which will hold samples. This attribute will be created with the source grid type if it does not exist, and with the source grid name if the name is empty
filter	an optional index filter
interrupter	an optional interrupter

Note

The target attribute may exist provided it can be cast to the SourceGridT ValueType

void rasterize ( const PointDataTreeOrGridT &  points, TransferT &  transfer, const FilterT &  filter = NullFilter(), InterrupterT *  interrupter = nullptr  )

inline

Perform potentially complex rasterization from a user defined transfer scheme.

A transfer scheme must be configured to call the provided

rasterize methods. See below for an example or PointRasterizeSDF.h/PointRasterizeTrilinear.h for implementations.

struct Transfer
{
    /// @return Returns the tree topology to loop over. This can be different
    ///   from the destination tree i.e. This can act as a mask.
    inline auto& topology();

    /// @brief  The maximum lookup range of this transfer scheme in index
    ///   space of the source points.
    /// @details The return value represent how far away from the destination
    ///   leaf node points should be accessed.
    /// @param origin  The leaf origin of the topology being accessed
    /// @param idx     The leaf index of the topology being accessed
    inline Int32 range(const Coord& origin, size_t idx) const;

    /// @brief  The initialize function, called on each leaf which has valid
    ///   topology to write to.
    /// @param origin  The leaf origin of the topology being accessed
    /// @param idx     The leaf index of the topology being accessed
    /// @param bounds  The active voxel bounds of the leaf
    inline void initialize(const Coord& origin, size_t idx, const CoordBBox& bounds);

    /// @brief  Run each time a point leaf is accessed. Typically this is
    ///  where attribute handles can be constructed
    /// @param leaf  The PointDataLeafNode which is being accessed.
    /// @return  Return true to continue rasterization, false to early exit
    ///   and skip the current leaf's contribution to the destination volume.
    inline bool startPointLeaf(const PointDataTree::LeafNodeType& leaf);

    /// @brief  The point stamp function. Each point which contributes to
    ///  the current leaf will call this function exactly once.
    /// @param ijk  The current voxel containing the point being rasterized.
    ///   May be outside the destination leaf node depending on the range()
    /// @param id   The point index being rasterized
    /// @param bounds  The active bounds of the leaf node.
    void rasterizePoint(const Coord& ijk,
                    const Index id,
                    const CoordBBox& bounds);

    /// @brief  Run each time a point leaf is finished with.
    /// @param leaf  The PointDataLeafNode which was being accessed.
    /// @return  Return true to continue rasterization, false to early exit
    ///   and stop rasterization to the destination leaf node.
    inline bool endPointLeaf(const PointDataTree::LeafNodeType& leaf);

    /// @brief  The finalization function for the given destination tree(s).
    /// @param origin  The leaf origin of the topology being accessed
    /// @param idx     The leaf index of the topology being accessed
    /// @return  Return true to stop, false to recursively rasterize
    inline bool finalize(const Coord& origin, size_t idx);
};

Below is a full example using the native components.

/// @brief Sum point distances into a target float tree
///   Note: Using TransformTransfer to handle different index spaces, and
///   VolumeTransfer for automatic buffer setup
struct MyTransfer :
    public TransformTransfer,
    public VolumeTransfer<FloatTree>
{
    MyTransfer(FloatGrid& dest, const PointDataGrid& source)
        : TransformTransfer(source.transform(), dest.transform())
        , VolumeTransfer(dest.tree())
        , mHandle(nullptr) {}

    MyTransfer(const MyTransfer& other)
        : TransformTransfer(other)
        , VolumeTransfer(other)
        , mHandle(nullptr) {}

    /// @brief Range in index space of the source points
    Int32 range(const Coord&, size_t) const { return Int32(1); }

    /// @brief Every time we start a new point leaf, init the position array.
    ///   Always return true as we don't skip any leaf nodes.
    bool startPointLeaf(const PointDataTree::LeafNodeType& leaf)
    {
        mHandle.reset(new AttributeHandle<Vec3f>(leaf.constAttributeArray("P"));
        return true;
    }

    /// @brief  For each point, compute its relative index space position in
    ///   the destination tree and sum the length of its distance
    void rasterizePoint(const Coord& ijk, const Index id, const CoordBBox& bounds)
    {
        Vec3d P = ijk.asVec3d() + Vec3d(this->mHandle->get(id));
        P = this->transformSourceToTarget(P); // TransformTransfer::transformSourceToTarget
        // for each active voxel, accumulate distance
        const auto* mask = this->mask(); // VolumeTransfer::mask
        for (auto& coord : bounds) {
            const Index voxel = FloatTree::LeafNodeType::coordToOffset(coord);
            if (!mask->isOn(voxel)) continue;
            Vec3d dist = coord.asVec3d() - P;
            this->buffer()[voxel] += dist.length(); // VolumeTransfer::buffer
        }
    }

    /// @brief Return true for endPointLeaf() to continue, false for finalize() so
    ///   we don't recurse.
    bool endPointLeaf(const PointDataTree::LeafNodeType&) { return true; }
    bool finalize(const Coord&, size_t) { return false; }

private:
    std::unique_ptr<AttributeHandle<Vec3f>> mHandle;
};

The method works by looping over a single Tree topology, looking up point data at a position relative to that topology and passing that data to a transfer scheme TransferT.

Note

Each thread receives a copy of the transfer scheme object.

Parameters

points	the point data grid to rasterize
transfer	the transfer scheme
filter	optional point filter
interrupter	optional interrupter

SdfT::Ptr rasterizeSmoothSpheres	(	const PointDataGridT &	points,
		const Real	radius,
		const Real	searchRadius,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Smoothed point distribution based sphere stamping with a uniform radius.

Rasterizes points into a level set using [Zhu Bridson 05] sphere stamping with a uniform radius. The radius and search radius parameters are given in world space units and are applied to every point to generate a fixed surface mask and consequent distance values. The search radius is each points points maximum contribution to the target level set. The search radius should always have a value equal to or larger than the point radius.

Warning

The width of the exterior half band may be smaller than the specified half band if the search radius is less than the equivalent world space halfband distance.

Parameters

points	the point data grid to rasterize
radius	the world space radius of every point
searchRadius	the maximum search distance of every point
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

The signed distance field.

SdfT::Ptr rasterizeSmoothSpheres	(	const PointDataGridT &	points,
		const std::string &	radius,
		const Real	radiusScale,
		const Real	searchRadius,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Smoothed point distribution based sphere stamping with a varying radius.

Rasterizes points into a level set using [Zhu Bridson 05] sphere stamping with a variable radius. The radius string parameter expects a point attribute of type RadiusT to exist. The radiusScale parameter is multiplier for radius values held on the radius attribute. The searchRadius parameter remains a fixed size value which represents each points points maximum contribution to the target level set. The radius scale and search radius parameters are given in world space units and are applied to every point to generate a fixed surface mask and consequent distance values. The search radius should always have a value equal to or larger than the point radii.

Warning

The width of the exterior half band may be smaller than the specified half band if the search radius is less than the equivalent world space halfband distance.

Parameters

points	the point data grid to rasterize
radius	the attribute containing the world space radius
radiusScale	the scale applied to every world space radius value
searchRadius	the maximum search distance of every point
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

The signed distance field.

GridPtrVec rasterizeSmoothSpheres	(	const PointDataGridT &	points,
		const Real	radius,
		const Real	searchRadius,
		const std::vector< std::string > &	attributes,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Smoothed point distribution based sphere stamping with a uniform radius and closest point attribute transfer.

Rasterizes points into a level set using [Zhu Bridson 05] sphere stamping with a uniform radius. The radius and search radius parameters are given in world space units and are applied to every point to generate a fixed surface mask and consequent distance values. The search radius is each points points maximum contribution to the target level set. The search radius should always be larger than the point radius. Every voxel's closest point is used to transfer each attribute in the attributes parameter to a new grid of matching topology. The destination types of these grids is equal to the ValueConverter result of the attribute type applied to the PointDataGridT.

Note

The AttributeTypes template parameter should be a TypeList of the required or possible attributes types. i.e. TypeList<int, float, double>. A runtime error will be thrown if no equivalent type for a given attribute is found in the AttributeTypes TypeList.

Warning

The width of the exterior half band may be smaller than the specified half band if the search radius is less than the equivalent world space halfband distance.

Parameters

points	the point data grid to rasterize
radius	the world space radius of every point
searchRadius	the maximum search distance of every point
attributes	list of attributes to transfer
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

A vector of grids. The signed distance field is guaranteed to be first and at the type specified by SdfT. Successive grids are the closest point attribute grids. These grids are guaranteed to have a topology and transform equal to the surface.

GridPtrVec rasterizeSmoothSpheres	(	const PointDataGridT &	points,
		const std::string &	radius,
		const Real	radiusScale,
		const Real	searchRadius,
		const std::vector< std::string > &	attributes,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Smoothed point distribution based sphere stamping with a varying radius and closest point attribute transfer.

Rasterizes points into a level set using [Zhu Bridson 05] sphere stamping with a variable radius. The radius string parameter expects a point attribute of type RadiusT to exist. The radiusScale parameter is multiplier for radius values held on the radius attribute. The searchRadius parameter remains a fixed size value which represents each points points maximum contribution to the target level set. The radius scale and search radius parameters are given in world space units and are applied to every point to generate a fixed surface mask and consequent distance values. The search radius should always have a value equal to or larger than the point radii. Every voxel's closest point is used to transfer each attribute in the attributes parameter to a new grid of matching topology. The destination types of these grids is equal to the ValueConverter result of the attribute type applied to the PointDataGridT.

Note

The AttributeTypes template parameter should be a TypeList of the required or possible attributes types. i.e. TypeList<int, float, double>. A runtime error will be thrown if no equivalent type for a given attribute is found in the AttributeTypes TypeList.

Warning

The width of the exterior half band may be smaller than the specified half band if the search radius is less than the equivalent world space halfband distance.

Parameters

points	the point data grid to rasterize
radius	the attribute containing the world space radius
radiusScale	the scale applied to every world space radius value
searchRadius	the maximum search distance of every point
attributes	list of attributes to transfer
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

A vector of grids. The signed distance field is guaranteed to be first and at the type specified by SdfT. Successive grids are the closest point attribute grids. These grids are guaranteed to have a topology and transform equal to the surface.

SdfT::Ptr rasterizeSpheres	(	const PointDataGridT &	points,
		const Real	radius,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Narrow band sphere stamping with a uniform radius.

Rasterizes points into a level set using basic sphere stamping with a uniform radius. The radius parameter is given in world space units and is applied to every point to generate a fixed surface mask and consequent distance values.

Parameters

points	the point data grid to rasterize
radius	the world space radius of every point
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

The signed distance field.

SdfT::Ptr rasterizeSpheres	(	const PointDataGridT &	points,
		const std::string &	radius,
		const Real	scale = 1.0,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Narrow band sphere stamping with a varying radius.

Rasterizes points into a level set using basic sphere stamping with a variable radius. The radius string parameter expects a point attribute of type RadiusT to exist.

Parameters

points	the point data grid to rasterize
radius	the name of the radius attribute
scale	an optional scale to apply to each per point radius
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

The signed distance field.

GridPtrVec rasterizeSpheres	(	const PointDataGridT &	points,
		const Real	radius,
		const std::vector< std::string > &	attributes,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Narrow band sphere stamping with a uniform radius and closest point attribute transfer.

Rasterizes points into a level set using basic sphere stamping with a uniform radius. The radius parameter is given in world space units and is applied to every point to generate a fixed surface mask and consequent distance values. Every voxel's closest point is used to transfer each attribute in the attributes parameter to a new grid of matching topology. The destination types of these grids is equal to the ValueConverter result of the attribute type applied to the PointDataGridT.

Note

The AttributeTypes template parameter should be a TypeList of the required or possible attributes types. i.e. TypeList<int, float, double>. A runtime error will be thrown if no equivalent type for a given attribute is found in the AttributeTypes TypeList.

Parameters

points	the point data grid to rasterize
radius	the world space radius of every point
attributes	list of attributes to transfer
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

A vector of grids. The signed distance field is guaranteed to be first and at the type specified by SdfT. Successive grids are the closest point attribute grids. These grids are guaranteed to have a topology and transform equal to the surface.

GridPtrVec rasterizeSpheres	(	const PointDataGridT &	points,
		const std::string &	radius,
		const std::vector< std::string > &	attributes,
		const Real	scale = 1.0,
		const Real	halfband = LEVEL_SET_HALF_WIDTH,
		math::Transform::Ptr	transform = nullptr,
		const FilterT &	filter = NullFilter(),
		InterrupterT *	interrupter = nullptr
	)

Narrow band sphere stamping with a varying radius and closest point attribute transfer.

Rasterizes points into a level set using basic sphere stamping with a variable radius. The radius string parameter expects a point attribute of type RadiusT to exist. Every voxel's closest point is used to transfer each attribute in the attributes parameter to a new grid of matching topology. The destination types of these grids is equal to the ValueConverter result of the attribute type applied to the PointDataGridT.

Note

The AttributeTypes template parameter should be a TypeList of the required or possible attributes types. i.e. TypeList<int, float, double>. A runtime error will be thrown if no equivalent type for a given attribute is found in the AttributeTypes TypeList.

Parameters

points	the point data grid to rasterize
radius	the name of the radius attribute
attributes	list of attributes to transfer
scale	scale to apply to each per point radius
halfband	the half band width
transform	the target transform for the surface
filter	a filter to apply to points
interrupter	optional interrupter

Returns

A vector of grids. The signed distance field is guaranteed to be first and at the type specified by SdfT. Successive grids are the closest point attribute grids. These grids are guaranteed to have a topology and transform equal to the surface.

auto rasterizeTrilinear ( const PointDataTreeT &  points, const std::string &  attribute, const FilterT &  filter = NullFilter()  )

inline

Perform weighted trilinear rasterization of all points within a voxel. This method takes and returns a tree i.e. ignores grid transformations.

Accumulates values and weights according to a simple 0-1-0 weighted hat function. This algorithm is an exact inverse of a trilinear interpolation and thus a key method used in PIC/FLIP style simulations. Returns a tree of the same precision as the input source attribute, but may be of a different math type depending on the value of the Staggered template attribute. If Staggered is true, this method produces values at each voxels negative faces, causing scalar attributes to produce math::Vec3<ValueT> tree types. The result Tree type is equal to: TrilinearTraits<ValueT, Staggered>::template TreeT<PointDataTreeT>

Template Parameters

Staggered	whether to perform a staggered or collocated rasterization
ValueT	the value type of the point attribute to rasterize

Parameters

points	the point tree to be rasterized
attribute	the name of the attribute to rasterize. Must be a scalar or Vec3 attribute.
filter	an optional point filter to use

void renameAttribute ( PointDataTreeT &  tree, const Name &  oldName, const Name &  newName  )

inline

Rename an attribute in a VDB tree.

Parameters

tree	the PointDataTree.
oldName	the old attribute name to rename from.
newName	the new attribute name to rename to.

Note

newName must not already exist and must not be a group attribute.

void renameAttributes ( PointDataTreeT &  tree, const std::vector< Name > &  oldNames, const std::vector< Name > &  newNames  )

inline

Rename attributes in a VDB tree.

Parameters

tree	the PointDataTree.
oldNames	a list of old attribute names to rename from.
newNames	a list of new attribute names to rename to.

Note

Number of oldNames must match the number of newNames.

Duplicate names and renaming group attributes are not allowed.

PointDataGridT::Ptr replicate	(	const PointDataGridT &	source,
		const Index	multiplier,
		const std::vector< std::string > &	attributes,
		const std::string &	scaleAttribute = "",
		const std::string &	replicationIndex = ""
	)

Replicates points provided in a source grid into a new grid, transfering and creating attributes found in a provided attribute vector. If an attribute doesn't exist, it is ignored. Position is always replicated, leaving the new points exactly over the top of the source points.

Note

The position attribute must exist

Parameters

source	The source grid to replicate points from
multiplier	The base number of points to replicate per point
attributes	Attributes to transfer to the new grid
scaleAttribute	A scale float attribute which multiplies the base multiplier to vary the point count per point.
replicationIndex	When provided, creates a replication attribute of the given name which holds the replication index. This can be subsequently used to modify the replicated points as a post process.

PointDataGridT::Ptr replicate	(	const PointDataGridT &	source,
		const Index	multiplier,
		const std::string &	scaleAttribute = "",
		const std::string &	replicationIndex = ""
	)

Replicates points provided in a source grid into a new grid, transfering and creating all attributes from the source grid. Position is always replicated, leaving the new points exactly over the top of the source points.

Note

The position attribute must exist

Parameters

source	The source grid to replicate points from
multiplier	The base number of points to replicate per point
scaleAttribute	A scale float attribute which multiplies the base multiplier to vary the point count per point.
replicationIndex	When provided, creates a replication attribute of the given name which holds the replication index. This can be subsequently used to modify the replicated points as a post process.

void sampleGrid ( size_t  order, PointDataGridT &  points, const SourceGridT &  sourceGrid, const Name &  targetAttribute, const FilterT &  filter = NullFilter(), const SamplerT &  sampler = SampleWithRounding(), InterrupterT *const  interrupter = nullptr, const bool  threaded = true  )

inline

Performs sampling and conversion from a VDB grid onto a VDB Points attribute.

Parameters

order	the sampling order - 0 = closest-point, 1 = trilinear, 2 = triquadratic
points	the PointDataGrid whose points will be sampled on to
sourceGrid	VDB grid which will be sampled
targetAttribute	a target attribute on the points which will hold samples. This attribute will be created with the source grid type if it does not exist, and with the source grid name if the name is empty
filter	an optional index filter
sampler	handles sampling and conversion into the target attribute type, which by default this uses the SampleWithRounding struct.
interrupter	an optional interrupter
threaded	enable or disable threading (threading is enabled by default)

Note

The target attribute may exist provided it can be cast to the SourceGridT ValueType

void setGroup ( PointDataTreeT &  tree, const PointIndexTreeT &  indexTree, const std::vector< short > &  membership, const Name &  group, const bool  remove = false  )

inline

Sets group membership from a PointIndexTree-ordered vector.

Parameters

tree	the PointDataTree.
indexTree	the PointIndexTree.
membership	1 if the point is in the group, 0 otherwise.
group	the name of the group.
remove	if true also perform removal of points from the group.

Note

vector<bool> is not thread-safe on concurrent write, so use vector<short> instead

void setGroup ( PointDataTreeT &  tree, const Name &  group, const bool  member = true  )

inline

Sets membership for the specified group for all points (on/off).

Parameters

tree	the PointDataTree.
group	the name of the group.
member	true / false for membership of the group.

void setGroupByFilter ( PointDataTreeT &  tree, const Name &  group, const FilterT &  filter  )

inline

Sets group membership based on a provided filter.

Parameters

tree	the PointDataTree.
group	the name of the group.
filter	filter data that is used to create a per-leaf filter

void openvdb::v12_0::points::setGroupByRandomPercentage ( PointDataTreeT &  tree, const Name &  group, const float  percentage = 10.0f, const unsigned int  seed = 0  )

inline

void openvdb::v12_0::points::setGroupByRandomTarget ( PointDataTreeT &  tree, const Name &  group, const Index64  targetPoints, const unsigned int  seed = 0  )

inline

void setStreamingMode ( PointDataTreeT &  tree, bool  on = true  )

inline

Toggle the streaming mode on all attributes in the tree to collapse the attributes after deconstructing a bound AttributeHandle to each array. This results in better memory efficiency when the data is streamed into another data structure (typically for rendering).

Parameters

tree	the PointDataTree.
on	true to enable streaming

Note

Multiple threads cannot safely access the same AttributeArray when using streaming.

PointDataGridT::Ptr uniformPointScatter ( const GridT &  grid, const Index64  count, const unsigned int  seed = 0, const float  spread = 1.0f, InterrupterT *  interrupter = nullptr  )

inline

The free functions depend on the following class:

The InterrupterT template argument below refers to any class with the following interface:

 class Interrupter {
   ...
 public:
   void start(const char* name = nullptr) // called when computations begin
   void end()                             // called when computations end
   bool wasInterrupted(int percent=-1)    // return true to break computation
};

Note

If no template argument is provided for this InterrupterT the util::NullInterrupter is used which implies that all interrupter calls are no-ops (i.e. incurs no computational overhead). Uniformly scatter a total amount of points in active regions

Parameters

grid	A source grid. The resulting PointDataGrid will copy this grids transform and scatter in its active voxelized topology.
count	The total number of points to scatter
seed	A seed for the RandGenT
spread	The spread of points as a scale from each voxels center. A value of 1.0f indicates points can be placed anywhere within the voxel, where as a value of 0.0f will force all points to be created exactly at the centers of each voxel.
interrupter	An optional interrupter

Note

returns the scattered PointDataGrid