Functions
void	knn (raft::handle_t &handle, std::vector< Matrix::Data< int64_t > * > out_I, std::vector< Matrix::floatData_t > out_D, std::vector< Matrix::floatData_t > &idx_data, Matrix::PartDescriptor &idx_desc, std::vector< Matrix::floatData_t * > &query_data, Matrix::PartDescriptor &query_desc, bool rowMajorIndex, bool rowMajorQuery, int k, size_t batch_size, bool verbose)

void	knn_classify (raft::handle_t &handle, std::vector< Matrix::Data< int > * > out, std::vector< std::vector< float >> probas, std::vector< Matrix::floatData_t > &idx_data, Matrix::PartDescriptor &idx_desc, std::vector< Matrix::floatData_t * > &query_data, Matrix::PartDescriptor &query_desc, std::vector< std::vector< int * >> &y, std::vector< int * > &uniq_labels, std::vector< int > &n_unique, bool rowMajorIndex=false, bool rowMajorQuery=false, bool probas_only=false, int k=10, size_t batch_size=1<< 15, bool verbose=false)

void	knn_regress (raft::handle_t &handle, std::vector< Matrix::Data< float > * > out, std::vector< Matrix::floatData_t > &idx_data, Matrix::PartDescriptor &idx_desc, std::vector< Matrix::floatData_t * > &query_data, Matrix::PartDescriptor &query_desc, std::vector< std::vector< float * >> &y, bool rowMajorIndex, bool rowMajorQuery, int k, int n_outputs, size_t batch_size, bool verbose)

Function Documentation

◆ knn()

void ML::KNN::opg::knn	(	raft::handle_t &	handle,
		std::vector< Matrix::Data< int64_t > * > *	out_I,
		std::vector< Matrix::floatData_t * > *	out_D,
		std::vector< Matrix::floatData_t * > &	idx_data,
		Matrix::PartDescriptor &	idx_desc,
		std::vector< Matrix::floatData_t * > &	query_data,
		Matrix::PartDescriptor &	query_desc,
		bool	rowMajorIndex,
		bool	rowMajorQuery,
		int	k,
		size_t	batch_size,
		bool	verbose
	)

Performs a multi-node multi-GPU KNN.

Parameters

[in]	handle	the raft::handle_t to use for managing resources
[out]	out_I	vector of output index partitions. size should match the number of local input partitions.
[out]	out_D	vector of output distance partitions. size should match the number of local input partitions.
[in]	idx_data	vector of local indices to query
[in]	idx_desc	describes how the index partitions are distributed across the ranks.
[in]	query_data	vector of local query partitions
[in]	query_desc	describes how the query partitions are distributed across the cluster.
[in]	rowMajorIndex	boolean indicating whether the index is row major.
[in]	rowMajorQuery	boolean indicating whether the query is row major.
[in]	k	the number of neighbors to query
[in]	batch_size	the max number of rows to broadcast at a time
[in]	verbose	print extra logging info

void ML::KNN::opg::knn_classify	(	raft::handle_t &	handle,
		std::vector< Matrix::Data< int > * > *	out,
		std::vector< std::vector< float * >> *	probas,
		std::vector< Matrix::floatData_t * > &	idx_data,
		Matrix::PartDescriptor &	idx_desc,
		std::vector< Matrix::floatData_t * > &	query_data,
		Matrix::PartDescriptor &	query_desc,
		std::vector< std::vector< int * >> &	y,
		std::vector< int * > &	uniq_labels,
		std::vector< int > &	n_unique,
		bool	rowMajorIndex = `false`,
		bool	rowMajorQuery = `false`,
		bool	probas_only = `false`,
		int	k = `10`,
		size_t	batch_size = `1<< 15`,
		bool	verbose = `false`
	)

Performs a multi-node multi-GPU KNN classify.

Parameters

[in]	handle	the raft::handle_t to use for managing resources
[out]	out	vector of output labels partitions. size should match the number of local input partitions.
[in]	probas	(optional) pointer to a vector containing arrays of probabilities
[in]	idx_data	vector of local indices to query
[in]	idx_desc	describes how the index partitions are distributed across the ranks.
[in]	query_data	vector of local query partitions
[in]	query_desc	describes how the query partitions are distributed across the cluster.
[in]	y	vector of vector of label arrays. for multilabel classification, each element in the vector is a different "output" array of labels corresponding to the i'th output. size should match the number of local input partitions.
[in]	uniq_labels	vector of the sorted unique labels for each array in y
[in]	n_unique	vector of sizes for each array in uniq_labels
[in]	rowMajorIndex	boolean indicating whether the index is row major.
[in]	rowMajorQuery	boolean indicating whether the query is row major.
[in]	probas_only	return probas instead of performing complete knn_classify
[in]	k	the number of neighbors to query
[in]	batch_size	the max number of rows to broadcast at a time
[in]	verbose	print extra logging info

void ML::KNN::opg::knn_regress	(	raft::handle_t &	handle,
		std::vector< Matrix::Data< float > * > *	out,
		std::vector< Matrix::floatData_t * > &	idx_data,
		Matrix::PartDescriptor &	idx_desc,
		std::vector< Matrix::floatData_t * > &	query_data,
		Matrix::PartDescriptor &	query_desc,
		std::vector< std::vector< float * >> &	y,
		bool	rowMajorIndex,
		bool	rowMajorQuery,
		int	k,
		int	n_outputs,
		size_t	batch_size,
		bool	verbose
	)

Performs a multi-node multi-GPU KNN regress.

Parameters

[in]	handle	the raft::handle_t to use for managing resources
[out]	out	vector of output partitions. size should match the number of local input partitions.
[in]	idx_data	vector of local indices to query
[in]	idx_desc	describes how the index partitions are distributed across the ranks.
[in]	query_data	vector of local query partitions
[in]	query_desc	describes how the query partitions are distributed across the cluster.
[in]	y	vector of vector of output arrays. for multi-output regression, each element in the vector is a different "output" array corresponding to the i'th output. size should match the number of local input partitions.
[in]	rowMajorIndex	boolean indicating whether the index is row major.
[in]	rowMajorQuery	boolean indicating whether the query is row major.
[in]	k	the number of neighbors to query
[in]	n_outputs	number of outputs
[in]	batch_size	the max number of rows to broadcast at a time
[in]	verbose	print extra logging info