Build and query | Is it right? Measure recall | Tune at query time | Which index? | The kNN graph of the data itself | Metrics and precision | Threads | Saving and loading
Package: annsearchR 0.1.0
Gregor Lueg
annsearchR: Fast Approximate Nearest Neighbour Search
R interface to the Rust crate 'ann-search-rs'. Build an exact or approximate nearest neighbour index once, keep it in memory behind an 'R6' object and query it as often as you like. Index types include hierarchical navigable small world graphs (HNSW), 'Annoy' random projection trees, inverted file indices (IVF), nearest neighbour descent (NNDescent) and 'Vamana' graphs. Multi-threaded via the Rust crate 'Rayon', with 32-bit or 64-bit floating point storage. Methods are described in Malkov and Yashunin (2020) <doi:10.1109/TPAMI.2018.2889473> and Dong et al. (2011) <doi:10.1145/1963405.1963487>.
Authors:
annsearchR_0.1.0.tar.gz
annsearchR_0.1.0.zip(r-4.7-x86_64)annsearchR_0.1.0.zip(r-4.7-arm64)annsearchR_0.1.0.zip(r-4.6-x86_64)annsearchR_0.1.0.zip(r-4.6-arm64)
annsearchR_0.1.0.tar.gz(r-4.7-arm64)annsearchR_0.1.0.tar.gz(r-4.7-x86_64)annsearchR_0.1.0.tar.gz(r-4.6-arm64)annsearchR_0.1.0.tar.gz(r-4.6-x86_64)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
annsearchR/json (API)
| # Install 'annsearchR' in R: |
| install.packages('annsearchR', repos = c('https://cran.r-universe.dev', 'https://cloud.r-project.org')) |
Bug tracker:https://github.com/gregorlueg/annsearchr/issues
Pkgdown/docs site:https://gregorlueg.github.io
Last updated from:1e855569eb. Checks:9 OK, 1 FAIL. Indexed: yes.
| Target | Result | Time | Files | Syslog |
|---|---|---|---|---|
| linux-devel-arm64 | OK | 263 | ||
| linux-devel-x86_64 | OK | 284 | ||
| source / vignettes | OK | 571 | ||
| linux-release-arm64 | OK | 240 | ||
| linux-release-x86_64 | OK | 244 | ||
| windows-devel-arm64 | OK | 579 | ||
| windows-devel-x86_64 | OK | 540 | ||
| windows-release-arm64 | OK | 526 | ||
| windows-release-x86_64 | OK | 484 | ||
| wasm-release | FAIL | 155 |
Exports:ann_get_threadsann_set_threadsAnnIndexAnnoyIndexBallTreeIndexExhaustiveIndexgenerate_cell_embeddingsgenerate_clustered_datagenerate_correlated_datagenerate_low_rank_dataHnswIndexIvfIndexKdTreeIndexKmknnIndexknn_recallload_ann_indexLshIndexNNDescentIndexNsgIndexRnnDescentIndexSoarIndexsubsample_queriesVamanaIndex
Readme and manuals
Help Manual
| Help page | Topics |
|---|---|
| Number of threads used for index builds and queries | ann_get_threads |
| Set the number of threads for index builds and queries | ann_set_threads |
| Base class for all nearest neighbour indices | AnnIndex |
| Annoy index | AnnoyIndex |
| Ball tree index | BallTreeIndex |
| Exhaustive (brute-force) index | ExhaustiveIndex |
| Generate synthetic cell embeddings | generate_cell_embeddings |
| Generate clustered synthetic data | generate_clustered_data |
| Generate correlated synthetic data | generate_correlated_data |
| Generate low-rank synthetic data | generate_low_rank_data |
| HNSW index | HnswIndex |
| IVF index | IvfIndex |
| kd forest index | KdTreeIndex |
| kMkNN index | KmknnIndex |
| Recall of an approximate kNN result | knn_recall |
| Load an index written by $save() | load_ann_index |
| LSH index | LshIndex |
| NN-Descent index | NNDescentIndex |
| NSG index | NsgIndex |
| Query an index | predict.AnnIndex |
| Relative NN-Descent index | RnnDescentIndex |
| SOAR index | SoarIndex |
| Subsample queries from a dataset | subsample_queries |
| Vamana index | VamanaIndex |