Modern columnar data format for ML and LLMs implemented in Rust. Convert from parquet in 2 lines of code for 100x faster random access, vector index, and data versioning. Compatible with Pandas, DuckDB, Polars, Pyarrow, and PyTorch with more integrations coming..
We have some basic merge insert functionality in lance today. Other databases have more sophisticated capabilities. Perhaps the most extensive and well documented is the "merge into" from SQL server. This issue will keep track of the various things that aren't yet implemented in our version.
[x] #2610
[ ] Allow update columns to be projection statements
This will allow for things like `when_matched().update({"count": "source.count + target.count"})
[ ] Allow updates to be applied to "when not matched by source" rows
This allows for updating target rows that don't match anything in the source (today we can just delete them)
We have some basic merge insert functionality in lance today. Other databases have more sophisticated capabilities. Perhaps the most extensive and well documented is the "merge into" from SQL server. This issue will keep track of the various things that aren't yet implemented in our version.