Data Bubbles / simMetrics
Implementation of "Data Bubbles: Quality Preserving Performance Boosting for Hierarchical Clustering" in Rust (Paper @ https://www.dbs.ifi.lmu.de/Publikationen/Papers/DataBubbles-cameraReady.pdf), and several molecular similarity metrics.
Followers
No followers yet
Ship Your Project
Get ready!
Once you ship this you can't edit the description of the project, but you'll be able to add more devlogs and re-ship it as you add new features!
Timeline
I worked on a significant amount of this project before summer of making, but I only have 2 hours from summer.
Today, I ran test clustering on 1000 molecules from the PubChem database and from that found several bugs that I fixed. Namely:
- Wrong bit similarity calculations for distance
- Overflow errors and array initialization errors in computing data bubbles.
I fixed these today, and will attempt to cluster larger amounts of molecules next.
Clusters are saved under tests/clusters.json, code in tests/genClusters.rs.
I'm also planning on figuring out ways of visualizing these high-dimensional clusters.
