Feature: Add Document Retrieval with Metadata Filtering similar to Chroma Vector Store #264
Conversation
…to default port used in tests
…ure/search_by_metadata
dishaprakash
left a comment
dishaprakash
left a comment
There was a problem hiding this comment.
Can we also add some examples related to this new feature into the vector store how to notebook?
|
@dishaprakash done |
|
@gRedHeadphone as we add new features to PGVectorStore, lets also update the how to guide for PGVectorstore on the langchain website (ref PR - langchain-ai/langchain#32549) |
| results = result_map.fetchall() | ||
| return bool(len(results) == 1) | ||
|
|
||
| async def aget( |
There was a problem hiding this comment.
Can we also support an ids input and just call the get_by_ids method?
There was a problem hiding this comment.
but then other parameters won't work. How about updating filters with ids filter?
|
|
||
| async def aget( | ||
| self, | ||
| filter: Optional[dict] = None, |
There was a problem hiding this comment.
We most likely should use "where" to match Chroma
| limit: Optional[int] = None, | ||
| offset: Optional[int] = None, | ||
| **kwargs: Any, | ||
| ) -> list[Document]: |
There was a problem hiding this comment.
This method should return A dict with the keys "ids", "embeddings", "metadatas", "documents" to match the Chroma implementation
There was a problem hiding this comment.
so should I also add "include" parameter?
3 participants
original PR: #231
On top of it adding limit and offset parameter to get methods similar to chroma vector store