feature: removed pandas dataframe dependency for similary_search when using DuckDB as vector store #30445
Conversation
… using DuckDB as vector store
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎ 1 Skipped Deployment
|
| return ids | ||
|
|
||
| def similarity_search( | ||
| def similarity_search_pd( |
There was a problem hiding this comment.
We should ensure that the behavior of similarity_search is identical to what it was previously, and not keep the pandas stuff around IMO.
There was a problem hiding this comment.
That I did verify with a test example, with text and metadata filled in. Doesn't seem to be breaking anything. However, I'd like you to verify as well, just to be doubly sure.
ccurme
left a comment
ccurme
left a comment
There was a problem hiding this comment.
Could you run the tests in tests/integration_tests/vectorstores/test_duckdb.py? I'm seeing some of them fail in this branch.
|
Great catch! I've updated the code now. I missed a small, but critical case when metadata is None, but is passed to json.loads() as such which caused the issue. Rectified now. |
2 participants
PR title: "community: Removes pandas dependency for using DuckDB for similarity search"
PR message:
similarity_search_pd, while the new one is atsimilarity_searchand requires no code changes. Return format remains the same.langchain_community.vectorstores.DuckDBrequires pandas - add a helpful (error) message #29933 and update on PR feature: added warning when duckdb is used as a vectorstore without pandas #30435