Package index • rnndescent

Main interface

Find approximate neighbors or build and query an index.

rnnd_knn(): Find approximate nearest neighbors
rnnd_build(): Build approximate nearest neighbors index and neighbor graph
rnnd_query(): Query an index for approximate nearest neighbors

Find or query neighbors using brute force, random selection, nearest neighbor descent, or random projection forests.

brute_force_knn(): Find exact nearest neighbors by brute force
brute_force_knn_query(): Query exact nearest neighbors by brute force
random_knn(): Find nearest neighbors by random selection
random_knn_query(): Query nearest neighbors by random selection
nnd_knn(): Find nearest neighbors using nearest neighbor descent
rpf_knn(): Find nearest neighbors using a random projection forest
rpf_build(): Create a random projection forest nearest neighbor index
rpf_knn_query(): Query a random projection forest index for nearest neighbors
rpf_filter(): Keep the best trees in a random projection forest

Prepare and query search graphs, merge neighbor graphs, and assess their properties.

prepare_search_graph(): Convert a nearest neighbor graph into a search graph
graph_knn_query(): Query a search graph for nearest neighbors
merge_knn(): Merge multiple approximate nearest neighbors graphs
k_occur(): Quantify hubness of a nearest neighbor graph
neighbor_overlap(): Overlap between the indices of two nearest neighbor graphs
truncate_knn() sort_knn() remove_self_knn() self_first_knn(): Manipulate nearest neighbor graphs
knn_to_sparse(): Convert a neighbor graph to a sparse matrix
knn_distances(): Calculate distances for fixed neighbor indices