A comprehensive knowledge base extracted directly from the official Macaulay2 source code and documentation packages (M2/Macaulay2/packages). Designed to be ingested into Vector Databases (e.g., ChromaDB, FAISS) for Retrieval-Augmented Generation (RAG).
Total Chunks: 15,711
Size: ~45MB
Code Density: ~85% of chunks contain executable Macaulay2 examples.
Source: Parsed from 2,208 .m2 files using a custom AST/Regex-aware parser… See the full description on the dataset page:
https://huggingface.co/datasets/frupniew/macaulay2-rag-chunks.
