HMLP: High-performance Machine Learning Primitives
|
This the splitter used in the randomized tree. More...
#include <gofmm.hpp>
Public Member Functions | |
randomsplit (SPDMATRIX &K) | |
vector< vector< size_t > > | operator() (vector< size_t > &gids) const |
Public Attributes | |
SPDMATRIX * | Kptr = NULL |
DistanceMetric | metric = ANGLE_DISTANCE |
This the splitter used in the randomized tree.
end struct centersplit
|
inline |
overload with the operator
Randomly select two points p and q.
Compute all pairwise distances.
SPDMATRIX* hmlp::gofmm::randomsplit< SPDMATRIX, N_SPLIT, T >::Kptr = NULL |
closure
DistanceMetric hmlp::gofmm::randomsplit< SPDMATRIX, N_SPLIT, T >::metric = ANGLE_DISTANCE |
(default) using angle distance from the Gram vector space