Bruteforce

The bruteforce method is running the KNN algorithm. It performs an extensive search, and in contrast to ANN methods produces an exact result.

#include <raft/neighbors/bruteforce.h>

Index

typedef cuvsBruteForceIndex *cuvsBruteForceIndex_t

cuvsError_t cuvsBruteForceIndexCreate(cuvsBruteForceIndex_t *index)

Allocate BRUTEFORCE index.

Parameters:

index – [in] cuvsBruteForceIndex_t to allocate

Returns:

cuvsError_t

cuvsError_t cuvsBruteForceIndexDestroy(cuvsBruteForceIndex_t index)

De-allocate BRUTEFORCE index.

Parameters:

index – [in] cuvsBruteForceIndex_t to de-allocate

struct cuvsBruteForceIndex

#include <brute_force.h>

Struct to hold address of cuvs::neighbors::brute_force::index and its active trained dtype.

Index build

cuvsError_t cuvsBruteForceBuild(cuvsResources_t res, DLManagedTensor *dataset, enum DistanceType metric, float metric_arg, cuvsBruteForceIndex_t index)

Build a BRUTEFORCE index with a DLManagedTensor which has underlying DLDeviceType equal to kDLCUDA, kDLCUDAHost, kDLCUDAManaged, or kDLCPU. Also, acceptable underlying types are:

1. kDLDataType.code == kDLFloat and kDLDataType.bits = 32

2. kDLDataType.code == kDLInt and kDLDataType.bits = 8

3. kDLDataType.code == kDLUInt and kDLDataType.bits = 8

#include <cuvs/core/c_api.h>
#include <cuvs/neighbors/brute_force.h>

// Create cuvsResources_t
cuvsResources_t res;
cuvsError_t res_create_status = cuvsResourcesCreate(&res);

// Assume a populated `DLManagedTensor` type here
DLManagedTensor dataset;

// Create BRUTEFORCE index
cuvsBruteForceIndex_t index;
cuvsError_t index_create_status = cuvsBruteForceIndexCreate(&index);

// Build the BRUTEFORCE Index
cuvsError_t build_status = cuvsBruteForceBuild(res, &dataset_tensor, L2Expanded, 0.f, index);

// de-allocate `index` and `res`
cuvsError_t index_destroy_status = cuvsBruteForceIndexDestroy(index);
cuvsError_t res_destroy_status = cuvsResourcesDestroy(res);

Parameters:

• res – [in] cuvsResources_t opaque C handle

• dataset – [in] DLManagedTensor* training dataset

• metric – [in] metric

• metric_arg – [in] metric_arg

• index – [out] cuvsBruteForceIndex_t Newly built BRUTEFORCE index

Returns:

cuvsError_t

Index search

cuvsError_t cuvsBruteForceSearch(cuvsResources_t res, cuvsBruteForceIndex_t index, DLManagedTensor *queries, DLManagedTensor *neighbors, DLManagedTensor *distances)

Search a BRUTEFORCE index with a DLManagedTensor which has underlying DLDeviceType equal to kDLCUDA, kDLCUDAHost, kDLCUDAManaged. It is also important to note that the BRUTEFORCE index must have been built with the same type of queries, such that index.dtype.code == queries.dl_tensor.dtype.code Types for input are:

1. queries: kDLDataType.code == kDLFloat and kDLDataType.bits = 32

2. neighbors: kDLDataType.code == kDLUInt and kDLDataType.bits = 32

3. distances: kDLDataType.code == kDLFloat and kDLDataType.bits = 32

#include <cuvs/core/c_api.h>
#include <cuvs/neighbors/brute_force.h>

// Create cuvsResources_t
cuvsResources_t res;
cuvsError_t res_create_status = cuvsResourcesCreate(&res);

// Assume a populated `DLManagedTensor` type here
DLManagedTensor dataset;
DLManagedTensor queries;
DLManagedTensor neighbors;

// Search the `index` built using `cuvsBruteForceBuild`
cuvsError_t search_status = cuvsBruteForceSearch(res, index, &queries, &neighbors, &distances);

// de-allocate `res`
cuvsError_t res_destroy_status = cuvsResourcesDestroy(res);

Parameters:

• res – [in] cuvsResources_t opaque C handle

• index – [in] cuvsBruteForceIndex which has been returned by cuvsBruteForceBuild

• queries – [in] DLManagedTensor* queries dataset to search

• neighbors – [out] DLManagedTensor* output k neighbors for queries

• distances – [out] DLManagedTensor* output k distances for queries