#include <nanoflann.hpp>

Public Attributes
IndexType	ptIdx = 0
	index of the stored data point
Dimension	divfeat = 0
	splitting axis at this node
bool	deleted = false
	this node's point is tombstoned
bool	treeDeleted = false
	whole subtree lazily tombstoned
INode *	child1 = nullptr
	"< split" child (also free-list link)
INode *	child2 = nullptr
	">= split" child
INode *	parent = nullptr
	parent (nullptr at the root)
Size	subtree_size = 0
	number of nodes in this subtree
Size	invalid_count = 0
	number of tombstoned nodes in subtree
BoundingBox	box
	AABB of all points (live+dead) in this subtree Cache of this node's own point coordinates, kept in-node to avoid the dataset_get() indirection on the hot query / insert / box paths. Only populated for a compile-time fixed DIM (kCacheCoords); for DIM=-1 it stays an empty vector and the code falls back to the dataset.
array_or_vector< DIM, ElementType >::type	pcoord

Detailed Description

template<typename Distance, class DatasetAdaptor, int32_t DIM = -1, typename IndexType = uint32_t>
struct nanoflann::KDTreeSingleIndexIncrementalAdaptor< Distance, DatasetAdaptor, DIM, IndexType >::INode

Augmented tree node: stores a single data point plus the maintenance metadata. Children pointers are nullptr at the leaves.

The documentation for this struct was generated from the following file:

include/nanoflann.hpp

Public Attributes

Detailed Description