test_qdrant_mmr_search() — langchain Function Reference

Source Code

libs/partners/qdrant/tests/integration_tests/qdrant_vector_store/test_mmr.py lines 26–76

def test_qdrant_mmr_search(
    location: str,
    content_payload_key: str,
    metadata_payload_key: str,
    retrieval_mode: RetrievalMode,
    vector_name: str,
) -> None:
    """Test end to end construction and MRR search."""
    filter_ = models.Filter(
        must=[
            models.FieldCondition(
                key=f"{metadata_payload_key}.page",
                match=models.MatchValue(
                    value=2,
                ),
            ),
        ],
    )

    texts = ["foo", "bar", "baz"]
    metadatas = [{"page": i} for i in range(len(texts))]
    docsearch = QdrantVectorStore.from_texts(
        texts,
        ConsistentFakeEmbeddings(),
        metadatas=metadatas,
        content_payload_key=content_payload_key,
        metadata_payload_key=metadata_payload_key,
        location=location,
        retrieval_mode=retrieval_mode,
        vector_name=vector_name,
        distance=models.Distance.EUCLID,
        sparse_embedding=ConsistentFakeSparseEmbeddings(),
    )
    output = docsearch.max_marginal_relevance_search(
        "foo", k=2, fetch_k=3, lambda_mult=0.0
    )
    assert_documents_equals(
        output,
        [
            Document(page_content="foo", metadata={"page": 0}),
            Document(page_content="bar", metadata={"page": 1}),
        ],
    )

    output = docsearch.max_marginal_relevance_search(
        "foo", k=2, fetch_k=3, lambda_mult=0.0, filter=filter_
    )
    assert_documents_equals(
        output,
        [Document(page_content="baz", metadata={"page": 2})],
    )