jafar::sams::Kdtree Class Reference

---

Detailed Description

Implementation of Kdtree: main class

Definition at line 99 of file kdtree.hpp.

#include <kdtree.hpp>

List of all members.

Public Member Functions
	Kdtree (jblas::mat &data_in, bool sort_out=true, bool rearrange_in=true, int dim_in=-1)
	~Kdtree ()
	destructor
void	n_nearest_brute_force (jblas::vec &qv, int nn, Kdtree_result_vector &result)
void	n_nearest (jblas::vec &qv, int nn, Kdtree_result_vector &result)
	search for n nearest to a given query vector 'qv'.
void	n_nearest_around_point (int idxin, int correltime, int nn, Kdtree_result_vector &result)
void	r_nearest (jblas::vec &qv, double r2, Kdtree_result_vector &result)
void	r_nearest_around_point (int idxin, int correltime, double r2, Kdtree_result_vector &result)
int	r_count (jblas::vec &qv, double r2)
	count number of neighbors within square distance r2.
int	r_count_around_point (int idxin, int correltime, double r2)
	like r_count, c
Public Attributes
const jblas::mat &	the_data
const int	N
	number of data points
int	dim
	dimension of the Kdtree
bool	sort_results
	USERS set to 'true'.
const bool	rearrange
	are we rearranging?
Private Member Functions
void	set_data (jblas::mat &din)
void	build_tree ()
Kdtree_node *	build_tree_for_range (int l, int u, Kdtree_node *parent)
void	select_on_coordinate (int c, int k, int l, int u)
int	select_on_coordinate_value (int c, double alpha, int l, int u)
void	spread_in_coordinate (int c, int l, int u, interval &interv)
Private Attributes
Kdtree_node *	root
const jblas::mat *	data
vector< int >	ind
jblas::mat	rearranged_data
Static Private Attributes
static const int	bucketsize = 12
Friends
class	Kdtree_node
class	searchrecord

---

Constructor & Destructor Documentation

jafar::sams::Kdtree::Kdtree	(	jblas::mat &	data_in,
		bool	sort_out = true,
		bool	rearrange_in = true,
		int	dim_in = -1
	)

constructor, passing in a (m,n) data matrix, aka n m-dimensional vectors.

constructor, has optional 'dim_in' feature, to use only first 'dim_in' components for definition of nearest neighbors.

---

Member Function Documentation

void jafar::sams::Kdtree::n_nearest_around_point	(	int	idxin,
		int	correltime,
		int	nn,
		Kdtree_result_vector &	result
	)

search for 'nn' nearest to point [idxin] of the input data, excluding neighbors within correltime

void jafar::sams::Kdtree::n_nearest_brute_force	(	jblas::vec &	qv,
		int	nn,
		Kdtree_result_vector &	result
	)

search for n nearest to a given query vector 'qv' usin exhaustive slow search. For debugging, usually.

void jafar::sams::Kdtree::r_nearest	(	jblas::vec &	qv,
		double	r2,
		Kdtree_result_vector &	result
	)

search for all neighbors in ball of size (square Euclidean distance) r2. Return number of neighbors in 'result.size()',

void jafar::sams::Kdtree::r_nearest_around_point	(	int	idxin,
		int	correltime,
		double	r2,
		Kdtree_result_vector &	result
	)

like 'r_nearest', but around existing point, with decorrelation interval.

---

Member Data Documentation

const jblas::mat& jafar::sams::Kdtree::the_data

"the_data" is a reference to the underlying jblas matrix of the data to be included in the tree.

NOTE: this structure does *NOT* own the storage underlying this. Hence, it would be a very bad idea to change the underlying data during use of the search facilities of this tree. Also, the user must deallocate the memory underlying it.

Definition at line 109 of file kdtree.hpp.

---

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

• modules/sams/include/sams/kdtree.hpp