HybridSearchConfig#

class langchain_postgres.v2.hybrid_search_config.HybridSearchConfig(tsv_column: str | None = '', tsv_lang: str | None = 'pg_catalog.english', fts_query: str | None = '', fusion_function: ~typing.Callable[[~typing.Sequence[~sqlalchemy.engine.row.RowMapping], ~typing.Sequence[~sqlalchemy.engine.row.RowMapping], ~typing.Any], ~typing.Sequence[~typing.Any]] = weighted_sum_ranking>, fusion_function_parameters: dict[str, ~typing.Any] = , primary_top_k: int = 4, secondary_top_k: int = 4, index_name: str = 'langchain_tsv_index', index_type: str = 'GIN')[source]#

AlloyDB Vector Store Hybrid Search Config.

Queries might be slow if the hybrid search column does not exist. For best hybrid search performance, consider creating a TSV column and adding GIN index.

Attributes

fts_query

index_name

index_type

primary_top_k

secondary_top_k

tsv_column

tsv_lang

Methods

__init__([tsv_column, tsv_lang, fts_query, ...])

fusion_function(secondary_search_results[, ...])

Ranks documents using a weighted sum of scores from two sources.
Parameters:

  • tsv_column (str | None)

  • tsv_lang (str | None)

  • fts_query (str | None)

  • fusion_function (Callable[[Sequence[RowMapping], Sequence[RowMapping], Any], Sequence[Any]])

  • fusion_function_parameters (dict[str, Any])

  • primary_top_k (int)

  • secondary_top_k (int)

  • index_name (str)

  • index_type (str)

__init__(tsv_column: str | None = '', tsv_lang: str | None = 'pg_catalog.english', fts_query: str | None = '', fusion_function: ~typing.Callable[[~typing.Sequence[~sqlalchemy.engine.row.RowMapping], ~typing.Sequence[~sqlalchemy.engine.row.RowMapping], ~typing.Any], ~typing.Sequence[~typing.Any]] = weighted_sum_ranking>, fusion_function_parameters: dict[str, ~typing.Any] = , primary_top_k: int = 4, secondary_top_k: int = 4, index_name: str = 'langchain_tsv_index', index_type: str = 'GIN') → None#
Parameters:

  • tsv_column (str | None)

  • tsv_lang (str | None)

  • fts_query (str | None)

  • fusion_function (Callable[[Sequence[RowMapping], Sequence[RowMapping], Any], Sequence[Any]])

  • fusion_function_parameters (dict[str, Any])

  • primary_top_k (int)

  • secondary_top_k (int)

  • index_name (str)

  • index_type (str)

Return type:

None

fusion_function(
secondary_search_results: Sequence[RowMapping],
primary_results_weight: float = 0.5,
secondary_results_weight: float = 0.5,
fetch_top_k: int = 4,
) → Sequence[dict[str, Any]]#

Ranks documents using a weighted sum of scores from two sources.

Parameters:

  • primary_search_results (Sequence[RowMapping]) – A list of (document, distance) tuples from the primary search.

  • secondary_search_results (Sequence[RowMapping]) – A list of (document, distance) tuples from the secondary search.

  • primary_results_weight (float) – The weight for the primary source’s scores. Defaults to 0.5.

  • secondary_results_weight (float) – The weight for the secondary source’s scores. Defaults to 0.5.

  • fetch_top_k (int) – The number of documents to fetch after merging the results. Defaults to 4.

Returns:

A list of (document, distance) tuples, sorted by weighted_score in descending order.

Return type:

Sequence[dict[str, Any]]