LMDiskANN.jl
LMDiskANN.jl: An Implementation of the Low Memory Disk Approximate Nearest Neighbors Search Algorithm - Published in JOSS (2025)
Science Score: 93.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
✓DOI references
Found 9 DOI reference(s) in README and JOSS metadata -
✓Academic publication links
Links to: joss.theoj.org -
○Committers with academic emails
-
○Institutional organization owner
-
✓JOSS paper metadata
Published in Journal of Open Source Software
Keywords
Repository
Julia Implementation of Low Memory Disk ANN (LM-DiskANN)
Basic Info
- Host: GitHub
- Owner: mantzaris
- License: mit
- Language: Julia
- Default Branch: main
- Homepage: https://mantzaris.github.io/LMDiskANN.jl
- Size: 757 KB
Statistics
- Stars: 6
- Watchers: 1
- Forks: 1
- Open Issues: 0
- Releases: 2
Topics
Metadata Files
README.md
<!-- -->
LMDiskANN.jl
Julia Implementation of Low Memory Disk ANN (LM-DiskANN)
LM-DiskANN is a lightweight library for approximate nearest‐neighbor (ANN) indexing on disk. It creates a graph structure over vector embeddings, storing them in memory‐mapped files to keep the in‐memory footprint low. It also supports optional LevelDB databases for user‐key ↔ numeric ID lookups, making it easy to associate each embedding with a custom string ID.
Key Features
- Disk‐Resident: Vectors and adjacency lists are stored in memory‐mapped files, reducing RAM usage.
- Graph‐Based Search: Leverages a BFS expansion (
EF_SEARCH) for approximate neighbor lookups. - Insert & Delete: Dynamically insert or remove embeddings without needing to rebuild the index.
- User Keys: Link a string key (e.g.
"image123") to your internal node ID; retrieve or delete by either integer ID or key.
Quick Start Example
Install via: (@v1.6) pkg> add https://github.com/mantzaris/LMDiskANN.jl
```julia using LMDiskANN using Random
create an index with dimension = 5
index = createindex("myindex", 5)
insert a random vector
v1 = rand(Float32, 5) (key1, id1) = ann_insert!(index, v1) # returns (autoKey, 1)
insert another vector with a custom string key
v2 = rand(Float32, 5) (mykey, id2) = ann_insert!(index, v2; key="myvec")
search for v1
results = search(index, v1, topk=3) println("Results for v1 => ", results)
retrieve embeddings
retrievedv1 = getembeddingfromid(index, id1) retrievedv2 = getembeddingfromkey(index, "myvec")
delete by ID or key
anndelete!(index, id1) anndelete!(index, "myvec") ```
citing this work
Mantzaris, A. V., (2025). LMDiskANN.jl: An Implementation of the Low Memory Disk Approximate Nearest Neighbors Search Algorithm. Journal of Open Source Software, 10(110), 8199, https://doi.org/10.21105/joss.08199
@article{Mantzaris2025, doi = {10.21105/joss.08199}, url = {https://doi.org/10.21105/joss.08199}, year = {2025}, publisher = {The Open Journal}, volume = {10}, number = {110}, pages = {8199}, author = {Alexander V. Mantzaris}, title = {LMDiskANN.jl: An Implementation of the Low Memory Disk Approximate Nearest Neighbors Search Algorithm}, journal = {Journal of Open Source Software} }
original paper introducing the LM-DiskANN algorithm
"LM-diskann: Low memory footprint in disk-native dynamic graph-based ann indexing." Pan, Yu, Jianxin Sun, and Hongfeng Yu, 2023 IEEE International Conference on Big Data (BigData). IEEE, 2023.
Owner
- Name: a.v.mantzaris
- Login: mantzaris
- Kind: user
- Location: USA
- Twitter: avmantzaris
- Repositories: 34
- Profile: https://github.com/mantzaris
Excited about the future of technology. Happy to participate in shaping that future through theory and practice.
JOSS Publication
LMDiskANN.jl: An Implementation of the Low Memory Disk Approximate Nearest Neighbors Search Algorithm
Authors
Department of Statistics and Data Science, University of Central Florida (UCF), USA
Editor
Mehmet Hakan Satman
Tags
ANN Embeddings Search Vector DBGitHub Events
Total
- Create event: 4
- Commit comment event: 8
- Issues event: 8
- Release event: 1
- Watch event: 3
- Delete event: 4
- Issue comment event: 10
- Push event: 55
- Pull request event: 1
- Fork event: 1
Last Year
- Create event: 4
- Commit comment event: 8
- Issues event: 8
- Release event: 1
- Watch event: 3
- Delete event: 4
- Issue comment event: 10
- Push event: 55
- Pull request event: 1
- Fork event: 1
Committers
Last synced: 8 months ago
Top Committers
| Name | Commits | |
|---|---|---|
| mantzaris | a****s@g****m | 44 |
| Mehmet Hakan Satman | m****n@g****m | 2 |
Issues and Pull Requests
Last synced: 8 months ago
All Time
- Total issues: 5
- Total pull requests: 1
- Average time to close issues: about 20 hours
- Average time to close pull requests: about 7 hours
- Total issue authors: 2
- Total pull request authors: 1
- Average comments per issue: 4.4
- Average comments per pull request: 0.0
- Merged pull requests: 1
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 5
- Pull requests: 1
- Average time to close issues: about 20 hours
- Average time to close pull requests: about 7 hours
- Issue authors: 2
- Pull request authors: 1
- Average comments per issue: 4.4
- Average comments per pull request: 0.0
- Merged pull requests: 1
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
- findmyway (4)
- JuliaTagBot (1)
Pull Request Authors
- jbytecode (2)
Top Labels
Issue Labels
Pull Request Labels
Packages
- Total packages: 1
- Total downloads: unknown
- Total dependent packages: 0
- Total dependent repositories: 0
- Total versions: 1
juliahub.com: LMDiskANN
Julia Implementation of Low Memory Disk ANN (LM-DiskANN)
- Homepage: https://mantzaris.github.io/LMDiskANN.jl
- Documentation: https://docs.juliahub.com/General/LMDiskANN/stable/
- License: MIT
-
Latest release: 1.1.0
published 8 months ago
Rankings
Dependencies
- actions/checkout v4 composite
- julia-actions/setup-julia v2 composite
- actions/checkout v4 composite
- julia-actions/cache v2 composite
- julia-actions/setup-julia v2 composite