46 #ifndef NANOFLANN_HPP_ 47 #define NANOFLANN_HPP_ 59 #if !defined(NOMINMAX) && (defined(_WIN32) || defined(_WIN32_) || defined(WIN32) || defined(_WIN64)) 73 #define NANOFLANN_VERSION 0x122 77 template <
typename DistanceType,
typename IndexType =
size_t,
typename CountType =
size_t>
90 inline void init(IndexType* indices_, DistanceType* dists_)
96 dists[
capacity-1] = (std::numeric_limits<DistanceType>::max)();
99 inline CountType
size()
const 110 inline void addPoint(DistanceType dist, IndexType index)
113 for (i=
count; i>0; --i) {
114 #ifdef NANOFLANN_FIRST_MATCH // If defined and two points have the same distance, the one with the lowest-index will be returned first. 117 if (
dists[i-1]>dist) {
143 template <
typename DistanceType,
typename IndexType =
size_t>
151 inline RadiusResultSet(DistanceType radius_, std::vector<std::pair<IndexType,DistanceType> >& indices_dists) : radius(radius_), m_indices_dists(indices_dists)
159 inline void clear() { m_indices_dists.clear(); }
161 inline size_t size()
const {
return m_indices_dists.size(); }
163 inline bool full()
const {
return true; }
165 inline void addPoint(DistanceType dist, IndexType index)
168 m_indices_dists.push_back(std::make_pair(index,dist));
171 inline DistanceType
worstDist()
const {
return radius; }
186 if (m_indices_dists.empty())
throw std::runtime_error(
"Cannot invoke RadiusResultSet::worst_item() on an empty list of results.");
187 typedef typename std::vector<std::pair<IndexType,DistanceType> >::const_iterator DistIt;
188 DistIt it = std::max_element(m_indices_dists.begin(), m_indices_dists.end());
197 template <
typename PairType>
198 inline bool operator()(
const PairType &p1,
const PairType &p2)
const {
199 return p1.second < p2.second;
211 fwrite(&value,
sizeof(value),
count, stream);
217 size_t size = value.size();
218 fwrite(&
size,
sizeof(
size_t), 1, stream);
219 fwrite(&value[0],
sizeof(T),
size, stream);
225 size_t read_cnt = fread(&value,
sizeof(value),
count, stream);
226 if (read_cnt !=
count) {
227 throw std::runtime_error(
"Cannot read from file");
236 size_t read_cnt = fread(&
size,
sizeof(
size_t), 1, stream);
238 throw std::runtime_error(
"Cannot read from file");
241 read_cnt = fread(&value[0],
sizeof(T),
size, stream);
242 if (read_cnt!=
size) {
243 throw std::runtime_error(
"Cannot read from file");
257 template<
class T,
class DataSource,
typename _DistanceType = T>
265 L1_Adaptor(
const DataSource &_data_source) : data_source(_data_source) { }
270 const T* last = a +
size;
271 const T* lastgroup = last - 3;
275 while (a < lastgroup) {
276 const DistanceType diff0 = std::abs(a[0] - data_source.kdtree_get_pt(b_idx,d++));
277 const DistanceType diff1 = std::abs(a[1] - data_source.kdtree_get_pt(b_idx,d++));
278 const DistanceType diff2 = std::abs(a[2] - data_source.kdtree_get_pt(b_idx,d++));
279 const DistanceType diff3 = std::abs(a[3] - data_source.kdtree_get_pt(b_idx,d++));
280 result += diff0 + diff1 + diff2 + diff3;
282 if ((worst_dist>0)&&(result>worst_dist)) {
288 result += std::abs( *a++ - data_source.kdtree_get_pt(b_idx,d++) );
293 template <
typename U,
typename V>
296 return std::abs(a-b);
305 template<
class T,
class DataSource,
typename _DistanceType = T>
313 L2_Adaptor(
const DataSource &_data_source) : data_source(_data_source) { }
318 const T* last = a +
size;
319 const T* lastgroup = last - 3;
323 while (a < lastgroup) {
324 const DistanceType diff0 = a[0] - data_source.kdtree_get_pt(b_idx,d++);
325 const DistanceType diff1 = a[1] - data_source.kdtree_get_pt(b_idx,d++);
326 const DistanceType diff2 = a[2] - data_source.kdtree_get_pt(b_idx,d++);
327 const DistanceType diff3 = a[3] - data_source.kdtree_get_pt(b_idx,d++);
328 result += diff0 * diff0 + diff1 * diff1 + diff2 * diff2 + diff3 * diff3;
330 if ((worst_dist>0)&&(result>worst_dist)) {
336 const DistanceType diff0 = *a++ - data_source.kdtree_get_pt(b_idx,d++);
337 result += diff0 * diff0;
342 template <
typename U,
typename V>
354 template<
class T,
class DataSource,
typename _DistanceType = T>
365 return data_source.kdtree_distance(a,b_idx,
size);
368 template <
typename U,
typename V>
377 template<
class T,
class DataSource>
384 template<
class T,
class DataSource>
391 template<
class T,
class DataSource>
406 leaf_max_size(_leaf_max_size)
416 SearchParams(
int checks_IGNORED_ = 32,
float eps_ = 0,
bool sorted_ =
true ) :
417 checks(checks_IGNORED_),
eps(eps_), sorted(sorted_) {}
436 template <
typename T>
439 T* mem =
static_cast<T*
>( ::malloc(
sizeof(T)*
count));
503 while (base !=
nullptr) {
504 void *prev = *(
static_cast<void**
>( base));
526 if (
size > remaining) {
528 wastedMemory += remaining;
535 void* m = ::malloc(blocksize);
537 fprintf(stderr,
"Failed to allocate memory.\n");
542 static_cast<void**
>(m)[0] = base;
548 remaining = blocksize -
sizeof(
void*) - shift;
549 loc = (
static_cast<char*
>(m) +
sizeof(
void*) + shift);
552 loc =
static_cast<char*
>(loc) +
size;
567 template <
typename T>
570 T* mem =
static_cast<T*
>(this->malloc(
sizeof(T)*
count));
606 template <
typename T, std::
size_t N>
628 #if !defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION) && !defined(BOOST_MSVC_STD_ITERATOR) && !defined(BOOST_NO_STD_ITERATOR_TRAITS) 631 #elif defined(_MSC_VER) && (_MSC_VER == 1300) && defined(BOOST_DINKUMWARE_STDLIB) && (BOOST_DINKUMWARE_STDLIB == 310) 660 static bool empty() {
return false; }
662 enum { static_size = N };
664 inline void resize(
const size_t nElements) {
if (nElements!=N)
throw std::logic_error(
"Try to change the size of a CArray."); }
668 const T*
data()
const {
return elems; }
677 inline void assign (
const T& value) {
for (
size_t i=0;i<N;i++) elems[i]=value; }
679 void assign (
const size_t n,
const T& value) { assert(N==n);
for (
size_t i=0;i<N;i++) elems[i]=value; }
688 template <
int DIM,
typename T>
694 template <
typename T>
739 template <
typename Distance,
class DatasetAdaptor,
int DIM = -1,
typename IndexType =
size_t>
832 dataset(inputData), index_params(
params), root_node(nullptr),
distance(inputData)
834 m_size = dataset.kdtree_get_point_count();
835 m_size_at_index_build =
m_size;
836 dim = dimensionality;
838 m_leaf_max_size =
params.leaf_max_size;
852 m_size_at_index_build = 0;
862 m_size_at_index_build =
m_size;
864 computeBoundingBox(root_bbox);
865 root_node = divideTree(0,
m_size, root_bbox );
873 return static_cast<size_t>(DIM>0 ? DIM : dim);
900 template <
typename RESULTSET>
907 throw std::runtime_error(
"[nanoflann] findNeighbors() called before building the index.");
908 float epsError = 1+searchParams.
eps;
911 dists.assign((DIM>0 ? DIM : dim) ,0);
913 searchLevel(result, vec, root_node, distsq,
dists, epsError);
914 return result.full();
928 resultSet.
init(out_indices, out_distances_sq);
930 return resultSet.
size();
948 const size_t nFound = radiusSearchCustomCallback(query_point,resultSet,searchParams);
959 template <
class SEARCH_CALLBACK>
962 this->findNeighbors(resultSet, query_point, searchParams);
963 return resultSet.size();
973 m_size = dataset.kdtree_get_point_count();
975 for (
size_t i = 0; i <
m_size; i++) vind[i] = i;
980 return dataset.kdtree_get_pt(idx,component);
988 save_tree(stream, tree->
child1);
991 save_tree(stream, tree->
child2);
1000 if (tree->
child1!=NULL) {
1001 load_tree(stream, tree->
child1);
1003 if (tree->
child2!=NULL) {
1004 load_tree(stream, tree->
child2);
1011 bbox.resize((DIM>0 ? DIM : dim));
1012 if (dataset.kdtree_get_bbox(bbox))
1018 const size_t N = dataset.kdtree_get_point_count();
1019 if (!N)
throw std::runtime_error(
"[nanoflann] computeBoundingBox() called but no data points found.");
1020 for (
int i=0; i<(DIM>0 ? DIM : dim); ++i) {
1022 bbox[i].high = dataset_get(0,i);
1024 for (
size_t k=1; k<N; ++k) {
1025 for (
int i=0; i<(DIM>0 ? DIM : dim); ++i) {
1026 if (dataset_get(k,i)<bbox[i].low) bbox[i].low = dataset_get(k,i);
1027 if (dataset_get(k,i)>bbox[i].high) bbox[i].high = dataset_get(k,i);
1046 if ( (right-left) <=
static_cast<IndexType
>(m_leaf_max_size) ) {
1049 node->node_type.lr.right = right;
1052 for (
int i=0; i<(DIM>0 ? DIM : dim); ++i) {
1053 bbox[i].low = dataset_get(vind[left],i);
1054 bbox[i].high = dataset_get(vind[left],i);
1056 for (IndexType k=left+1; k<right; ++k) {
1057 for (
int i=0; i<(DIM>0 ? DIM : dim); ++i) {
1058 if (bbox[i].low>dataset_get(vind[k],i)) bbox[i].low=dataset_get(vind[k],i);
1059 if (bbox[i].high<dataset_get(vind[k],i)) bbox[i].high=dataset_get(vind[k],i);
1067 middleSplit_(&vind[0]+left, right-left, idx, cutfeat, cutval, bbox);
1069 node->node_type.sub.divfeat = cutfeat;
1072 left_bbox[cutfeat].high = cutval;
1073 node->child1 = divideTree(left, left+idx, left_bbox);
1076 right_bbox[cutfeat].low = cutval;
1077 node->child2 = divideTree(left+idx, right, right_bbox);
1079 node->node_type.sub.divlow = left_bbox[cutfeat].high;
1080 node->node_type.sub.divhigh = right_bbox[cutfeat].low;
1082 for (
int i=0; i<(DIM>0 ? DIM : dim); ++i) {
1083 bbox[i].low = std::min(left_bbox[i].low, right_bbox[i].low);
1084 bbox[i].high = std::max(left_bbox[i].high, right_bbox[i].high);
1094 min_elem = dataset_get(ind[0],element);
1095 max_elem = dataset_get(ind[0],element);
1096 for (IndexType i=1; i<
count; ++i) {
1098 if (
val<min_elem) min_elem =
val;
1099 if (
val>max_elem) max_elem =
val;
1107 for (
int i=1; i<(DIM>0 ? DIM : dim); ++i) {
1109 if (span>max_span) {
1115 for (
int i=0; i<(DIM>0 ? DIM : dim); ++i) {
1117 if (span>(1-EPS)*max_span) {
1119 computeMinMax(ind,
count, cutfeat, min_elem, max_elem);
1121 if (spread>max_spread) {
1123 max_spread = spread;
1128 DistanceType split_val = (bbox[cutfeat].low+bbox[cutfeat].high)/2;
1130 computeMinMax(ind,
count, cutfeat, min_elem, max_elem);
1132 if (split_val<min_elem) cutval = min_elem;
1133 else if (split_val>max_elem) cutval = max_elem;
1134 else cutval = split_val;
1136 IndexType lim1, lim2;
1137 planeSplit(ind,
count, cutfeat, cutval, lim1, lim2);
1139 if (lim1>
count/2) index = lim1;
1140 else if (lim2<
count/2) index = lim2;
1141 else index =
count/2;
1158 IndexType right =
count-1;
1160 while (left<=right && dataset_get(ind[left],cutfeat)<cutval) ++left;
1161 while (right && left<=right && dataset_get(ind[right],cutfeat)>=cutval) --right;
1162 if (left>right || !right)
break;
1163 std::swap(ind[left], ind[right]);
1173 while (left<=right && dataset_get(ind[left],cutfeat)<=cutval) ++left;
1174 while (right && left<=right && dataset_get(ind[right],cutfeat)>cutval) --right;
1175 if (left>right || !right)
break;
1176 std::swap(ind[left], ind[right]);
1188 for (
int i = 0; i < (DIM>0 ? DIM : dim); ++i) {
1189 if (vec[i] < root_bbox[i].low) {
1190 dists[i] =
distance.accum_dist(vec[i], root_bbox[i].low, i);
1193 if (vec[i] > root_bbox[i].high) {
1194 dists[i] =
distance.accum_dist(vec[i], root_bbox[i].high, i);
1206 template <
class RESULTSET>
1211 if ((node->
child1 ==
nullptr)&&(node->
child2 ==
nullptr)) {
1214 for (IndexType i=node->
node_type.
lr.left; i<node->node_type.lr.right; ++i) {
1215 const IndexType index = vind[i];
1217 if (dist<worst_dist) {
1218 result_set.addPoint(dist,vind[i]);
1233 if ((diff1+diff2)<0) {
1234 bestChild = node->
child1;
1235 otherChild = node->
child2;
1239 bestChild = node->
child2;
1240 otherChild = node->
child1;
1245 searchLevel(result_set, vec, bestChild, mindistsq,
dists, epsError);
1248 mindistsq = mindistsq + cut_dist - dst;
1249 dists[idx] = cut_dist;
1250 if (mindistsq*epsError<=result_set.worstDist()) {
1251 searchLevel(result_set, vec, otherChild, mindistsq,
dists, epsError);
1268 save_tree(stream, root_node);
1282 load_tree(stream, root_node);
1312 typedef typename Distance::template traits<num_t,self_t>::distance_t
metric_t;
1321 if (dims!=dimensionality)
throw std::runtime_error(
"Error: 'dimensionality' must match column count in data matrix");
1322 if (DIM>0 && static_cast<int>(dims)!=DIM)
1323 throw std::runtime_error(
"Data set dimensionality does not match the 'DIM' template argument");
1343 inline void query(
const num_t *query_point,
const size_t num_closest,
IndexType *out_indices,
num_t *out_distances_sq,
const int = 10)
const 1346 resultSet.
init(out_indices, out_distances_sq);
1362 return m_data_matrix.rows();
1370 const num_t d= p1[i]-m_data_matrix.coeff(idx_p2,i);
1378 return m_data_matrix.coeff(idx,
IndexType(dim));
1384 template <
class BBOX>
An L2-metric KD-tree adaptor for working with data directly stored in an Eigen Matrix, without duplicating the data storage.
std::pair< IndexType, DistanceType > worst_item() const
Find the worst result (furtherest neighbor) without copying or sorting Pre-conditions: size() > 0...
DistanceType accum_dist(const U a, const V b, int) const
std::reverse_iterator< iterator > reverse_iterator
size_t size() const
Returns number of points in dataset.
bool kdtree_get_bbox(BBOX &) const
void freeIndex()
Frees the previously-built index.
void swap(CArray< T, N > &y)
const DataSource & data_source
Metaprogramming helper traits class for the L2 (Euclidean) metric.
std::vector< IndexType > vind
Array of indices to vectors in the dataset.
Manhattan distance functor (generic version, optimized for high-dimensionality data sets)...
void loadIndex(FILE *stream)
Loads a previous index from a binary file.
Squared Euclidean (L2) distance functor (suitable for low-dimensionality datasets, like 2D or 3D point clouds) Corresponding distance traits: nanoflann::metric_L2_Simple.
const DistanceType radius
size_t usedMemory() const
Computes the inde memory usage Returns: memory used by the index.
KDTreeEigenMatrixAdaptor< MatrixType, DIM, Distance > self_t
~PooledAllocator()
Destructor.
size_t m_size_at_index_build
Number of points in the dataset when the index was built.
std::ptrdiff_t difference_type
RadiusResultSet(DistanceType radius_, std::vector< std::pair< IndexType, DistanceType > > &indices_dists)
const KDTreeSingleIndexAdaptorParams index_params
static void rangecheck(size_type i)
reverse_iterator rbegin()
void init(IndexType *indices_, DistanceType *dists_)
IndexType right
Indices of points in leaf node.
const self_t & derived() const
const_reference at(size_type i) const
mrpt::vision::TStereoCalibParams params
DistanceType accum_dist(const U a, const V b, int) const
DistanceType operator()(const T *a, const size_t b_idx, size_t size) const
L2_Simple_Adaptor< T, DataSource > distance_t
float eps
search for eps-approximate neighbours (default: 0)
const_reference front() const
void planeSplit(IndexType *ind, const IndexType count, int cutfeat, DistanceType &cutval, IndexType &lim1, IndexType &lim2)
Subdivide the list of points by a plane perpendicular on axe corresponding to the 'cutfeat' dimension...
void resize(const size_t nElements)
This method has no effects in this class, but raises an exception if the expected size does not match...
L2_Adaptor< T, DataSource > distance_t
DistanceType computeInitialDistances(const ElementType *vec, distance_vector_t &dists) const
DistanceType operator()(const T *a, const size_t b_idx, size_t size, DistanceType worst_dist=-1) const
L1_Adaptor< T, DataSource > distance_t
void load_value(FILE *stream, T &value, size_t count=1)
Used to declare fixed-size arrays when DIM>0, dynamically-allocated vectors when DIM=-1.
const DataSource & data_source
bool operator()(const PairType &p1, const PairType &p2) const
PairType will be typically: std::pair<IndexType,DistanceType>
KDTreeSingleIndexAdaptorParams(size_t _leaf_max_size=10)
void searchLevel(RESULTSET &result_set, const ElementType *vec, const NodePtr node, DistanceType mindistsq, distance_vector_t &dists, const float epsError) const
Performs an exact search in the tree starting from a node.
KDTreeEigenMatrixAdaptor(const int dimensionality, const MatrixType &mat, const int leaf_max_size=10)
The kd-tree index for the user to call its methods as usual with any other FLANN index.
void init_vind()
Make sure the auxiliary list vind has the same size than the current dataset, and re-generate if size...
DistanceType accum_dist(const U a, const V b, int) const
bool findNeighbors(RESULTSET &result, const ElementType *vec, const SearchParams &searchParams) const
Find set of nearest neighbors to vec[0:dim-1].
L2_Adaptor(const DataSource &_data_source)
void addPoint(DistanceType dist, IndexType index)
void query(const num_t *query_point, const size_t num_closest, IndexType *out_indices, num_t *out_distances_sq, const int=10) const
Query for the num_closest closest points to a given point (entered as query_point[0:dim-1]).
array_or_vector_selector< DIM, DistanceType >::container_t distance_vector_t
Define "distance_vector_t" as a fixed-size or variable-size container depending on "DIM"...
static size_type max_size()
std::vector< T > container_t
bool sorted
only for radius search, require neighbours sorted by distance (default: true)
const DatasetAdaptor & dataset
The dataset used by this index.
operator "<" for std::sort()
reference operator[](size_type i)
Distance::DistanceType DistanceType
void middleSplit_(IndexType *ind, IndexType count, IndexType &index, int &cutfeat, DistanceType &cutval, const BoundingBox &bbox)
MatrixType::Index IndexType
DistanceType operator()(const T *a, const size_t b_idx, size_t size, DistanceType worst_dist=-1) const
std::vector< std::pair< IndexType, DistanceType > > & m_indices_dists
T * allocate(size_t count=1)
Allocates (using C's malloc) a generic type T.
Distance::ElementType ElementType
num_t kdtree_distance(const num_t *p1, const IndexType idx_p2, IndexType size) const
size_t radiusSearch(const ElementType *query_point, const DistanceType &radius, std::vector< std::pair< IndexType, DistanceType > > &IndicesDists, const SearchParams &searchParams) const
Find all the neighbors to query_point[0:dim-1] within a maximum radius.
reference at(size_type i)
~KDTreeEigenMatrixAdaptor()
KNNResultSet(CountType capacity_)
size_t radiusSearchCustomCallback(const ElementType *query_point, SEARCH_CALLBACK &resultSet, const SearchParams &searchParams=SearchParams()) const
Just like radiusSearch() but with a custom callback class for each point found in the radius of the q...
void saveIndex(FILE *stream)
Stores the index in a binary file.
int fprintf(FILE *fil, const char *format,...) noexcept MRPT_printf_format_check(2
An OS-independent version of fprintf.
A result-set class used when performing a radius based search.
ElementType dataset_get(size_t idx, int component) const
Helper accessor to the dataset points:
DistanceType worstDist() const
KDTreeSingleIndexAdaptor(const int dimensionality, const DatasetAdaptor &inputData, const KDTreeSingleIndexAdaptorParams ¶ms=KDTreeSingleIndexAdaptorParams())
KDTree constructor.
const_reverse_iterator rend() const
T * allocate(const size_t count=1)
Allocates (using this pool) a generic type T.
A STL container (as wrapper) for arrays of constant size defined at compile time (class imported from...
int checks
Ignored parameter (Kept for compatibility with the FLANN interface).
L1_Adaptor(const DataSource &_data_source)
void addPoint(DistanceType dist, IndexType index)
const_iterator end() const
const_reference back() const
void free_all()
Frees all allocated memory chunks.
const_iterator begin() const
size_t knnSearch(const ElementType *query_point, const size_t num_closest, IndexType *out_indices, DistanceType *out_distances_sq, const int=10) const
Find the "num_closest" nearest neighbors to the query_point[0:dim-1].
Distance::template traits< num_t, self_t >::distance_t metric_t
size_t veclen() const
Returns the length of each point in the dataset.
num_t kdtree_get_pt(const IndexType idx, int dim) const
_DistanceType DistanceType
const_reference operator[](size_type i) const
array_or_vector_selector< DIM, Interval >::container_t BoundingBox
Define "BoundingBox" as a fixed-size or variable-size container depending on "DIM".
Squared Euclidean distance functor (generic version, optimized for high-dimensionality data sets)...
PooledAllocator()
Default constructor.
NodePtr root_node
The KD-tree used to find neighbours.
const DataSource & data_source
void assign(const T &value)
Parameters (see README.md)
size_t m_size
Number of current poins in the dataset.
DistanceType worstDist() const
_DistanceType DistanceType
void computeMinMax(IndexType *ind, IndexType count, int element, ElementType &min_elem, ElementType &max_elem)
NodePtr divideTree(const IndexType left, const IndexType right, BoundingBox &bbox)
Create a tree node that subdivides the list of vecs from vind[first] to vind[last].
L2_Simple_Adaptor(const DataSource &_data_source)
void assign(const size_t n, const T &value)
void save_tree(FILE *stream, NodePtr tree)
int dim
Dimensionality of each data point.
struct nanoflann::KDTreeSingleIndexAdaptor::Node::@17::leaf lr
Search options for KDTreeSingleIndexAdaptor::findNeighbors()
std::reverse_iterator< const_iterator > const_reverse_iterator
const size_t WORDSIZE
Pooled storage allocator.
void computeBoundingBox(BoundingBox &bbox)
size_t kdtree_get_point_count() const
void save_value(FILE *stream, const T &value, size_t count=1)
void load_tree(FILE *stream, NodePtr &tree)
PooledAllocator pool
Pooled memory allocator.
const_iterator begin() const
const MatrixType & m_data_matrix
int divfeat
Dimension used for subdivision.
Metaprogramming helper traits class for the L1 (Manhattan) metric.
union nanoflann::KDTreeSingleIndexAdaptor::Node::@17 node_type
Union used because a node can be either a LEAF node or a non-leaf node, so both data fields are never...
const_reverse_iterator rbegin() const
KDTreeSingleIndexAdaptor< metric_t, self_t, DIM, IndexType > index_t
void clear()
Clear the contents of this container.
void * malloc(const size_t req_size)
Returns a pointer to a piece of new memory of the given size in bytes allocated from the pool...
SearchParams(int checks_IGNORED_=32, float eps_=0, bool sorted_=true)
Note: The first argument (checks_IGNORED_) is ignored, but kept for compatibility with the FLANN inte...
struct nanoflann::KDTreeSingleIndexAdaptor::Node::@17::nonleaf sub
Metaprogramming helper traits class for the L2_simple (Euclidean) metric.
size_t m_size
Number of elements accessed with write access so far.
_DistanceType DistanceType
void set_radius_and_clear(const DistanceType r)
Clears the result set and adjusts the search radius.
double distance(const TPoint2D &p1, const TPoint2D &p2)
Gets the distance between two points in a 2D space.
const_iterator end() const
CArray< T, DIM > container_t
void buildIndex()
Builds the index.
const T & const_reference
Node * child2
Child nodes (both=NULL mean its a leaf node)