I built a project that implemented intervaltree in Rust and exposed PyO3 bindings as a drop-in replacement for Python's native intervaltree. It is significantly faster, and I will be adding more features, such as AVL and red-black trees for balancing.
And with a little work you can even use them to map ranges of keys to values in a way that's reminiscent of interval trees — e.g. https://crates.io/crates/rangemap. (Disclosure: that's my crate.)
Nice! I was only suggesting considering BTrees because they also play nice with caches, instead of the more conventional binary tree balancing mechanisms.
Hi everyone
I wanted to share this project that I am currently building, where I am building a hierarchical graph structure for the documentations of different packages. The function discovery part in function calling is the trickiest part as just dense retrieval won't help a lot. Rather we can build all the documentation into a hierarchical graph, and based on the function definitions, we can jump to the nodes that are relevant and narrow down our search space. I am hoping to brainstorm more about it and currently, it supports sklearn, pandas, and OpenBB. Working on huggingface and Ray.