knncolle_kmknn::KmknnOptions< Index_, Data_, Distance_, KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ > Struct Template Reference

Options for KmknnBuilder construction. More...

#include <Kmknn.hpp>

Public Attributes
double	power = 0.5
std::shared_ptr< kmeans::Initialize< KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ > >	initialize_algorithm
std::shared_ptr< kmeans::Refine< KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ > >	refine_algorithm

Detailed Description

template<typename Index_, typename Data_, typename Distance_, typename KmeansIndex_ = Index_, typename KmeansData_ = Data_, typename KmeansCluster_ = Index_, typename KmeansFloat_ = Distance_, class KmeansMatrix_ = kmeans::SimpleMatrix<KmeansIndex_, KmeansData_>>
struct knncolle_kmknn::KmknnOptions< Index_, Data_, Distance_, KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ >

Options for KmknnBuilder construction.

This can also be created via the KmknnBuilder::Options typedef, which ensures consistency with the template parameters used in KmknnBuilder.

Template Parameters

Index_	Integer type for the observation indices.
Data_	Numeric type for the input and query data.
Distance_	Floating-point type for the distances.
KmeansIndex_	Integer type of the observation indices for kmeans.
KmeansData_	Numeric type of the input data for kmeans.
KmeansCluster_	Integer type of the cluster identities for kmeans.
KmeansFloat_	Floating-point type of the cluster centroids.
KmeansMatrix_	Class of the input data matrix for kmeans. This should satisfy the kmeans::Matrix interface, most typically a kmeans::SimpleMatrix. (Note that this is a different class from the knncolle::Matrix interface!)

Member Data Documentation

initialize_algorithm

template<typename Index_ , typename Data_ , typename Distance_ , typename KmeansIndex_ = Index_, typename KmeansData_ = Data_, typename KmeansCluster_ = Index_, typename KmeansFloat_ = Distance_, class KmeansMatrix_ = kmeans::SimpleMatrix<KmeansIndex_, KmeansData_>>

std::shared_ptr<kmeans::Initialize<KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_> > knncolle_kmknn::KmknnOptions< Index_, Data_, Distance_, KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ >::initialize_algorithm

Initialization method for k-means clustering. If NULL, defaults to kmeans::InitializeKmeanspp.

power

template<typename Index_ , typename Data_ , typename Distance_ , typename KmeansIndex_ = Index_, typename KmeansData_ = Data_, typename KmeansCluster_ = Index_, typename KmeansFloat_ = Distance_, class KmeansMatrix_ = kmeans::SimpleMatrix<KmeansIndex_, KmeansData_>>

double knncolle_kmknn::KmknnOptions< Index_, Data_, Distance_, KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ >::power = 0.5

Power of the number of observations, to define the number of cluster centers. By default, a square root is performed.

refine_algorithm

template<typename Index_ , typename Data_ , typename Distance_ , typename KmeansIndex_ = Index_, typename KmeansData_ = Data_, typename KmeansCluster_ = Index_, typename KmeansFloat_ = Distance_, class KmeansMatrix_ = kmeans::SimpleMatrix<KmeansIndex_, KmeansData_>>

std::shared_ptr<kmeans::Refine<KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_> > knncolle_kmknn::KmknnOptions< Index_, Data_, Distance_, KmeansIndex_, KmeansData_, KmeansCluster_, KmeansFloat_, KmeansMatrix_ >::refine_algorithm

Refinement method for k-means clustering. If NULL, defaults to kmeans::RefineHartiganWong.

---

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

• /github/workspace/include/knncolle_kmknn/Kmknn.hpp