InMemoryVectorstoreWithSearch Class — langchain Architecture

Architecture documentation for the InMemoryVectorstoreWithSearch class in test_multi_vector.py from the langchain codebase.

Class python

Entity Profile

Dependency Diagram

graph TD
  8d3a5c80_ca3c_3791_b15b_403681cc6589["InMemoryVectorstoreWithSearch"]
  6e491709_d60f_689d_8a1a_c760b54fd120["InMemoryVectorStore"]
  8d3a5c80_ca3c_3791_b15b_403681cc6589 -->|extends| 6e491709_d60f_689d_8a1a_c760b54fd120
  ec1b85d3_6e5d_33cc_5e60_268fb7b7b716["test_multi_vector.py"]
  8d3a5c80_ca3c_3791_b15b_403681cc6589 -->|defined in| ec1b85d3_6e5d_33cc_5e60_268fb7b7b716
  76234a63_3ce3_8aed_1a37_7d86c4868564["_identity_fn()"]
  8d3a5c80_ca3c_3791_b15b_403681cc6589 -->|method| 76234a63_3ce3_8aed_1a37_7d86c4868564
  2d9987eb_3297_ae3b_a879_f09052500366["_select_relevance_score_fn()"]
  8d3a5c80_ca3c_3791_b15b_403681cc6589 -->|method| 2d9987eb_3297_ae3b_a879_f09052500366
  8338cd8b_80db_db0a_805f_575dd8de2cf2["similarity_search()"]
  8d3a5c80_ca3c_3791_b15b_403681cc6589 -->|method| 8338cd8b_80db_db0a_805f_575dd8de2cf2
  5d810b0d_19ea_076f_3435_e71a46d0bc07["similarity_search_with_score()"]
  8d3a5c80_ca3c_3791_b15b_403681cc6589 -->|method| 5d810b0d_19ea_076f_3435_e71a46d0bc07

Relationship Graph

Source Code

libs/langchain/tests/unit_tests/retrievers/test_multi_vector.py lines 12–42

class InMemoryVectorstoreWithSearch(InMemoryVectorStore):
    @staticmethod
    def _identity_fn(score: float) -> float:
        return score

    def _select_relevance_score_fn(self) -> Callable[[float], float]:
        return self._identity_fn

    @override
    def similarity_search(
        self,
        query: str,
        k: int = 4,
        **kwargs: Any,
    ) -> list[Document]:
        res = self.store.get(query)
        if res is None:
            return []
        return [res]

    @override
    def similarity_search_with_score(
        self,
        query: str,
        k: int = 4,
        **kwargs: Any,
    ) -> list[tuple[Document, float]]:
        res = self.store.get(query)
        if res is None:
            return []
        return [(res, 0.8)]