STATUS: work in progress Fast Nearest-Neighbour searching. More...
#include "mathadd.h"#include "distances.h"#include "definitions.h"#include "optarg.h"
This graph shows which files directly or indirectly include this file:
Go to the source code of this file.
Data Structures | |
| struct | SearchTree |
Functions | |
| SearchTree * | nn_prepare (const double **X, int m, int N, OptArgList *optargs) |
| void | nn_search_k (const SearchTree *S, const double *x, int k, int *nn_idx, double *nn_dist) |
| void | nn_searck_k_slow (const double **X, int N, int m, const double *x, int k, int *nn_idx, double *nn_dist, OptArgList *optargs) |
Detailed Description
STATUS: work in progress Fast Nearest-Neighbour searching.
see
Merkwirth et al. Fast nearest-neighbor searching for nonlinear signal processing. Physical review E, Statistical physics, plasmas, fluids, and related interdisciplinary topics (2000) vol. 62 (2 Pt A) pp. 2089-97
Definition in file nnsearch.h.
Function Documentation
| SearchTree* nn_prepare | ( | const double ** | X, | |
| int | N, | |||
| int | m, | |||
| OptArgList * | optargs | |||
| ) |
Initialize the search tree for fast-nearest-neighbour searching.
- Parameters:
-
data matrix of m variables with N measurements each m dimensionality N number of measurements optargs may contain: - 'metric=void*' the distance metric used (default=euclidean)
- 'max_numel_terminal_node=int' the maximum number of elements in a terminal node (default=50)
- optional arguments for the called function (e.g. metric)
- Returns:
- the searchtree-struct used to look for nearest neighbours
Definition at line 38 of file nnsearch.c.
| void nn_search_k | ( | const SearchTree * | S, | |
| const double * | x, | |||
| int | k, | |||
| int * | nn_idx, | |||
| double * | nn_dist | |||
| ) |
Searching the k nearest neighbour of vector x in searchtree. This implementation uses a searchtree-based method.
- Parameters:
-
S the searchtree, previously initialized by nn_prepare x query point; must be of dimension m in searchtree S k number of neighbours to look for nn_idx memory for holding the indices for the k NN's nn_dist memory for holding distances for k NN's
Definition at line 147 of file nnsearch.c.
| void nn_searck_k_slow | ( | const double ** | X, | |
| int | N, | |||
| int | m, | |||
| const double * | x, | |||
| int | k, | |||
| int * | nn_idx, | |||
| double * | nn_dist, | |||
| OptArgList * | optargs | |||
| ) |
