PGVectorStore#

class langchain_postgres.v2.vectorstores.PGVectorStore( key: object, engine: PGEngine, vs: AsyncPGVectorStore, )[source]#

Postgres Vector Store class

PGVectorStore constructor. :param key: Prevent direct constructor usage. :type key: object :param engine: Connection pool engine for managing connections to Postgres database. :type engine: PGEngine :param vs: The async only VectorStore implementation :type vs: AsyncPGVectorStore

Raises:

Exception – If called directly by user.

Parameters:

key (object)
engine (PGEngine)
vs (AsyncPGVectorStore)

Attributes

embeddings

Access the query embedding object if available.

Methods

`__init__`(key, engine, vs)	PGVectorStore constructor.
`aadd_documents`(documents[, ids])	Embed documents and add to the table.
`aadd_embeddings`(texts, embeddings[, ...])	Add data along with embeddings to the table.
`aadd_texts`(texts[, metadatas, ids])	Embed texts and add to the table.
`aapply_vector_index`(index[, name, concurrently])	Create an index on the vector store table.
`add_documents`(documents[, ids])	Embed documents and add to the table.
`add_embeddings`(texts, embeddings[, ...])	Add data along with embeddings to the table.
`add_texts`(texts[, metadatas, ids])	Embed texts and add to the table.
`adelete`([ids])	Delete records from the table.
`adrop_vector_index`([index_name])	Drop the vector index.
`afrom_documents`(documents, embedding, ...[, ...])	Create an PGVectorStore instance from documents.
`afrom_texts`(texts, embedding, engine, table_name)	Create an PGVectorStore instance from texts.
`aget_by_ids`(ids)	Get documents by ids.
`ais_valid_index`([index_name])	Check if index exists in the table.
`amax_marginal_relevance_search`(query[, k, ...])	Return docs selected using the maximal marginal relevance.
`amax_marginal_relevance_search_by_vector`(...)	Return docs selected using the maximal marginal relevance.
`amax_marginal_relevance_search_with_score_by_vector`(...)	Return docs and distance scores selected using the maximal marginal relevance.
`apply_vector_index`(index[, name, concurrently])	Create an index on the vector store table.
`areindex`([index_name])	Re-index the vector store table.
`as_retriever`(**kwargs)	Return VectorStoreRetriever initialized from this VectorStore.
`asearch`(query, search_type, **kwargs)	Async return docs most similar to query using a specified search type.
`asimilarity_search`(query[, k, filter])	Return docs selected by similarity search on query.
`asimilarity_search_by_vector`(embedding[, k, ...])	Return docs selected by vector similarity search.
`asimilarity_search_with_relevance_scores`(query)	Async return docs and relevance scores in the range [0, 1].
`asimilarity_search_with_score`(query[, k, filter])	Return docs and distance scores selected by similarity search on query.
`asimilarity_search_with_score_by_vector`(...)	Return docs and distance scores selected by vector similarity search.
`create`(engine, embedding_service, table_name)	Create an PGVectorStore instance.
`create_sync`(engine, embedding_service, ...)	Create an PGVectorStore instance.
`delete`([ids])	Delete records from the table.
`drop_vector_index`([index_name])	Drop the vector index.
`from_documents`(documents, embedding, engine, ...)	Create an PGVectorStore instance from documents.
`from_texts`(texts, embedding, engine, table_name)	Create an PGVectorStore instance from texts.
`get_by_ids`(ids)	Get documents by ids.
`get_table_name`()
`is_valid_index`([index_name])	Check if index exists in the table.
`max_marginal_relevance_search`(query[, k, ...])	Return docs selected using the maximal marginal relevance.
`max_marginal_relevance_search_by_vector`(...)	Return docs selected using the maximal marginal relevance.
`max_marginal_relevance_search_with_score_by_vector`(...)	Return docs and distance scores selected using the maximal marginal relevance.
`reindex`([index_name])	Re-index the vector store table.
`search`(query, search_type, **kwargs)	Return docs most similar to query using a specified search type.
`similarity_search`(query[, k, filter])	Return docs selected by similarity search on query.
`similarity_search_by_vector`(embedding[, k, ...])	Return docs selected by vector similarity search.
`similarity_search_with_relevance_scores`(query)	Return docs and relevance scores in the range [0, 1].
`similarity_search_with_score`(query[, k, filter])	Return docs and distance scores selected by similarity search on query.
`similarity_search_with_score_by_vector`(embedding)	Return docs and distance scores selected by similarity search on vector.

__init__( key: object, engine: PGEngine, vs: AsyncPGVectorStore, )[source]#

PGVectorStore constructor. :param key: Prevent direct constructor usage. :type key: object :param engine: Connection pool engine for managing connections to Postgres database. :type engine: PGEngine :param vs: The async only VectorStore implementation :type vs: AsyncPGVectorStore

Raises:

Exception – If called directly by user.

Parameters:

key (object)
engine (PGEngine)
vs (AsyncPGVectorStore)

async aadd_documents(

documents: list[Document],

ids: list | None = None,

**kwargs: Any,

) → list[str][source]#

Embed documents and add to the table.

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Parameters:

documents (list[Document])
ids (list | None)
kwargs (Any)

Return type:

list[str]

async aadd_embeddings(

texts: Iterable[str],

embeddings: list[list[float]],

metadatas: list[dict] | None = None,

ids: list[str] | None = None,

**kwargs: Any,

) → list[str][source]#

Add data along with embeddings to the table.

Parameters:

texts (Iterable[str])
embeddings (list[list[float]])
metadatas (list[dict] | None)
ids (list[str] | None)
kwargs (Any)

Return type:

list[str]

async aadd_texts(

texts: Iterable[str],

metadatas: list[dict] | None = None,

ids: list | None = None,

**kwargs: Any,

) → list[str][source]#

Embed texts and add to the table.

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Parameters:

texts (Iterable[str])
metadatas (list[dict] | None)
ids (list | None)
kwargs (Any)

Return type:

list[str]

async aapply_vector_index( index: BaseIndex, name: str | None = None, concurrently: bool = False, ) → None[source]#

Create an index on the vector store table.

Parameters:

index (BaseIndex)
name (str | None)
concurrently (bool)

Return type:

None

add_documents(

documents: list[Document],

ids: list | None = None,

**kwargs: Any,

) → list[str][source]#

Embed documents and add to the table.

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Parameters:

documents (list[Document])
ids (list | None)
kwargs (Any)

Return type:

list[str]

add_embeddings(

texts: Iterable[str],

embeddings: list[list[float]],

metadatas: list[dict] | None = None,

ids: list[str] | None = None,

**kwargs: Any,

) → list[str][source]#

Add data along with embeddings to the table.

Parameters:

texts (Iterable[str])
embeddings (list[list[float]])
metadatas (list[dict] | None)
ids (list[str] | None)
kwargs (Any)

Return type:

list[str]

add_texts(

texts: Iterable[str],

metadatas: list[dict] | None = None,

ids: list | None = None,

**kwargs: Any,

) → list[str][source]#

Embed texts and add to the table.

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Parameters:

texts (Iterable[str])
metadatas (list[dict] | None)
ids (list | None)
kwargs (Any)

Return type:

list[str]

async adelete(

ids: list | None = None,

**kwargs: Any,

) → bool | None[source]#

Delete records from the table.

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Parameters:

ids (list | None)
kwargs (Any)

Return type:

bool | None

async adrop_vector_index( index_name: str | None = None, ) → None[source]#

Drop the vector index.

Parameters:: index_name (str | None)
Return type:: None

async classmethod afrom_documents(

documents: list[Document],

embedding: Embeddings,

engine: PGEngine,

table_name: str,

schema_name: str = 'public',

ids: list | None = None,

content_column: str = 'content',

embedding_column: str = 'embedding',

metadata_columns: list[str] | None = None,

ignore_metadata_columns: list[str] | None = None,

id_column: str = 'langchain_id',

metadata_json_column: str = 'langchain_metadata',

distance_strategy: DistanceStrategy = DistanceStrategy.COSINE_DISTANCE,

k: int = 4,

fetch_k: int = 20,

lambda_mult: float = 0.5,

index_query_options: QueryOptions | None = None,

**kwargs: Any,

) → PGVectorStore[source]#

Create an PGVectorStore instance from documents.

Parameters:

documents (list[Document]) – Documents to add to the vector store.
embedding (Embeddings) – Text embedding model to use.
engine (PGEngine) – Connection pool engine for managing connections to postgres database.
table_name (str) – Name of an existing table.
schema_name (str, optional) – Name of the database schema. Defaults to “public”.
ids (list | None) – (Optional[list]): List of IDs to add to table records. Defaults to None.
content_column (str, optional) – Column that represent a Document’s page_content. Defaults to “content”.
embedding_column (str, optional) – Column for embedding vectors. The embedding is generated from the document value. Defaults to “embedding”.
metadata_columns (list[str], optional) – Column(s) that represent a document’s metadata. Defaults to an empty list.
ignore_metadata_columns (Optional[list[str]], optional) – Column(s) to ignore in pre-existing tables for a document’s metadata. Can not be used with metadata_columns. Defaults to None.
id_column (str, optional) – Column that represents the Document’s id. Defaults to “langchain_id”.
metadata_json_column (str, optional) – Column to store metadata as JSON. Defaults to “langchain_metadata”.
distance_strategy (DistanceStrategy) – Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
k (int) – Number of Documents to return from search. Defaults to 4.
fetch_k (int) – Number of Documents to fetch to pass to MMR algorithm.
lambda_mult (float) – Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
index_query_options (QueryOptions) – Index query option.
kwargs (Any)

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Returns:

PGVectorStore

Return type:

PGVectorStore

async classmethod afrom_texts(

texts: list[str],

embedding: Embeddings,

engine: PGEngine,

table_name: str,

schema_name: str = 'public',

metadatas: list[dict] | None = None,

ids: list | None = None,

content_column: str = 'content',

embedding_column: str = 'embedding',

metadata_columns: list[str] | None = None,

ignore_metadata_columns: list[str] | None = None,

id_column: str = 'langchain_id',

metadata_json_column: str = 'langchain_metadata',

distance_strategy: DistanceStrategy = DistanceStrategy.COSINE_DISTANCE,

k: int = 4,

fetch_k: int = 20,

lambda_mult: float = 0.5,

index_query_options: QueryOptions | None = None,

**kwargs: Any,

) → PGVectorStore[source]#

Create an PGVectorStore instance from texts.

Parameters:

texts (list[str]) – Texts to add to the vector store.
embedding (Embeddings) – Text embedding model to use.
engine (PGEngine) – Connection pool engine for managing connections to postgres database.
table_name (str) – Name of an existing table.
schema_name (str, optional) – Name of the database schema. Defaults to “public”.
metadatas (Optional[list[dict]], optional) – List of metadatas to add to table records. Defaults to None.
ids (list | None) – (Optional[list]): List of IDs to add to table records. Defaults to None.
content_column (str, optional) – Column that represent a Document’s page_content. Defaults to “content”.
embedding_column (str, optional) – Column for embedding vectors. The embedding is generated from the document value. Defaults to “embedding”.
metadata_columns (list[str], optional) – Column(s) that represent a document’s metadata. Defaults to an empty list.
ignore_metadata_columns (Optional[list[str]], optional) – Column(s) to ignore in pre-existing tables for a document’s metadata. Can not be used with metadata_columns. Defaults to None.
id_column (str, optional) – Column that represents the Document’s id. Defaults to “langchain_id”.
metadata_json_column (str, optional) – Column to store metadata as JSON. Defaults to “langchain_metadata”.
distance_strategy (DistanceStrategy) – Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
k (int) – Number of Documents to return from search. Defaults to 4.
fetch_k (int) – Number of Documents to fetch to pass to MMR algorithm.
lambda_mult (float) – Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
index_query_options (QueryOptions) – Index query option.
kwargs (Any)

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Returns:

PGVectorStore

Return type:

PGVectorStore

async aget_by_ids( ids: Sequence[str], ) → list[Document][source]#

Get documents by ids.

Parameters:: ids (Sequence[str])
Return type:: list[Document]

async ais_valid_index( index_name: str | None = None, ) → bool[source]#

Check if index exists in the table.

Parameters:: index_name (str | None)
Return type:: bool

async amax_marginal_relevance_search(

query: str,

k: int | None = None,

fetch_k: int | None = None,

lambda_mult: float | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected using the maximal marginal relevance.

Parameters:

query (str)
k (int | None)
fetch_k (int | None)
lambda_mult (float | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

async amax_marginal_relevance_search_by_vector(

embedding: list[float],

k: int | None = None,

fetch_k: int | None = None,

lambda_mult: float | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected using the maximal marginal relevance.

Parameters:

embedding (list[float])
k (int | None)
fetch_k (int | None)
lambda_mult (float | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

async amax_marginal_relevance_search_with_score_by_vector(

embedding: list[float],

k: int | None = None,

fetch_k: int | None = None,

lambda_mult: float | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[tuple[Document, float]][source]#

Return docs and distance scores selected using the maximal marginal relevance.

Parameters:

embedding (list[float])
k (int | None)
fetch_k (int | None)
lambda_mult (float | None)
filter (dict | None)
kwargs (Any)

Return type:

list[tuple[Document, float]]

apply_vector_index( index: BaseIndex, name: str | None = None, concurrently: bool = False, ) → None[source]#

Create an index on the vector store table.

Parameters:

index (BaseIndex)
name (str | None)
concurrently (bool)

Return type:

None

async areindex( index_name: str | None = None, ) → None[source]#

Re-index the vector store table.

Parameters:: index_name (str | None)
Return type:: None

as_retriever(

**kwargs: Any,

) → VectorStoreRetriever#

Return VectorStoreRetriever initialized from this VectorStore.

Parameters:

**kwargs (Any) –

Keyword arguments to pass to the search function. Can include: search_type (Optional[str]): Defines the type of search that

the Retriever should perform. Can be “similarity” (default), “mmr”, or “similarity_score_threshold”.

search_kwargs (Optional[Dict]): Keyword arguments to pass to the

search function. Can include things like:

k: Amount of documents to return (Default: 4) score_threshold: Minimum relevance threshold

for similarity_score_threshold

fetch_k: Amount of documents to pass to MMR algorithm: (Default: 20)
lambda_mult: Diversity of results returned by MMR;: 1 for minimum diversity and 0 for maximum. (Default: 0.5)

filter: Filter by document metadata

Returns:

Retriever class for VectorStore.

Return type:

VectorStoreRetriever

Examples:

# Retrieve more documents with higher diversity
# Useful if your dataset has many similar documents
docsearch.as_retriever(
    search_type="mmr",
    search_kwargs={'k': 6, 'lambda_mult': 0.25}
)

# Fetch more documents for the MMR algorithm to consider
# But only return the top 5
docsearch.as_retriever(
    search_type="mmr",
    search_kwargs={'k': 5, 'fetch_k': 50}
)

# Only retrieve documents that have a relevance score
# Above a certain threshold
docsearch.as_retriever(
    search_type="similarity_score_threshold",
    search_kwargs={'score_threshold': 0.8}
)

# Only get the single most similar document from the dataset
docsearch.as_retriever(search_kwargs={'k': 1})

# Use a filter to only retrieve documents from a specific paper
docsearch.as_retriever(
    search_kwargs={'filter': {'paper_title':'GPT-4 Technical Report'}}
)

async asearch(

query: str,

search_type: str,

**kwargs: Any,

) → list[Document]#

Async return docs most similar to query using a specified search type.

Parameters:

query (str) – Input text.
search_type (str) – Type of search to perform. Can be “similarity”, “mmr”, or “similarity_score_threshold”.
**kwargs (Any) – Arguments to pass to the search method.

Returns:

List of Documents most similar to the query.

Raises:

ValueError – If search_type is not one of “similarity”, “mmr”, or “similarity_score_threshold”.

Return type:

list[Document]

async asimilarity_search(

query: str,

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected by similarity search on query.

Parameters:

query (str)
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

async asimilarity_search_by_vector(

embedding: list[float],

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected by vector similarity search.

Parameters:

embedding (list[float])
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

async asimilarity_search_with_relevance_scores(

query: str,

k: int = 4,

**kwargs: Any,

) → list[tuple[Document, float]]#

Async return docs and relevance scores in the range [0, 1].

0 is dissimilar, 1 is most similar.

Parameters:

query (str) – Input text.
k (int) – Number of Documents to return. Defaults to 4.
**kwargs (Any) –
kwargs to be passed to similarity search. Should include: score_threshold: Optional, a floating point value between 0 to 1 to

filter the resulting set of retrieved docs

Returns:

List of Tuples of (doc, similarity_score)

Return type:

list[tuple[Document, float]]

async asimilarity_search_with_score(

query: str,

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[tuple[Document, float]][source]#

Return docs and distance scores selected by similarity search on query.

Parameters:

query (str)
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[tuple[Document, float]]

async asimilarity_search_with_score_by_vector(

embedding: list[float],

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[tuple[Document, float]][source]#

Return docs and distance scores selected by vector similarity search.

Parameters:

embedding (list[float])
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[tuple[Document, float]]

async classmethod create( engine: PGEngine, embedding_service: Embeddings, table_name: str, schema_name: str = 'public', content_column: str = 'content', embedding_column: str = 'embedding', metadata_columns: list[str] | None = None, ignore_metadata_columns: list[str] | None = None, id_column: str = 'langchain_id', metadata_json_column: str | None = 'langchain_metadata', distance_strategy: DistanceStrategy = DistanceStrategy.COSINE_DISTANCE, k: int = 4, fetch_k: int = 20, lambda_mult: float = 0.5, index_query_options: QueryOptions | None = None, ) → PGVectorStore[source]#

Create an PGVectorStore instance.

Parameters:

engine (PGEngine) – Connection pool engine for managing connections to postgres database.
embedding_service (Embeddings) – Text embedding model to use.
table_name (str) – Name of an existing table.
schema_name (str, optional) – Name of the database schema. Defaults to “public”.
content_column (str) – Column that represent a Document’s page_content. Defaults to “content”.
embedding_column (str) – Column for embedding vectors. The embedding is generated from the document value. Defaults to “embedding”.
metadata_columns (list[str]) – Column(s) that represent a document’s metadata.
ignore_metadata_columns (list[str]) – Column(s) to ignore in pre-existing tables for a document’s metadata. Can not be used with metadata_columns. Defaults to None.
id_column (str) – Column that represents the Document’s id. Defaults to “langchain_id”.
metadata_json_column (str) – Column to store metadata as JSON. Defaults to “langchain_metadata”.
distance_strategy (DistanceStrategy) – Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
k (int) – Number of Documents to return from search. Defaults to 4.
fetch_k (int) – Number of Documents to fetch to pass to MMR algorithm.
lambda_mult (float) – Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
index_query_options (QueryOptions) – Index query option.

Returns:

PGVectorStore

Return type:

PGVectorStore

classmethod create_sync( engine: PGEngine, embedding_service: Embeddings, table_name: str, schema_name: str = 'public', content_column: str = 'content', embedding_column: str = 'embedding', metadata_columns: list[str] | None = None, ignore_metadata_columns: list[str] | None = None, id_column: str = 'langchain_id', metadata_json_column: str = 'langchain_metadata', distance_strategy: DistanceStrategy = DistanceStrategy.COSINE_DISTANCE, k: int = 4, fetch_k: int = 20, lambda_mult: float = 0.5, index_query_options: QueryOptions | None = None, ) → PGVectorStore[source]#

Create an PGVectorStore instance.

Parameters:

key (object) – Prevent direct constructor usage.
engine (PGEngine) – Connection pool engine for managing connections to postgres database.
embedding_service (Embeddings) – Text embedding model to use.
table_name (str) – Name of an existing table.
schema_name (str, optional) – Name of the database schema. Defaults to “public”.
content_column (str, optional) – Column that represent a Document’s page_content. Defaults to “content”.
embedding_column (str, optional) – Column for embedding vectors. The embedding is generated from the document value. Defaults to “embedding”.
metadata_columns (list[str], optional) – Column(s) that represent a document’s metadata. Defaults to None.
ignore_metadata_columns (Optional[list[str]]) – Column(s) to ignore in pre-existing tables for a document’s metadata. Can not be used with metadata_columns. Defaults to None.
id_column (str, optional) – Column that represents the Document’s id. Defaults to “langchain_id”.
metadata_json_column (str, optional) – Column to store metadata as JSON. Defaults to “langchain_metadata”.
distance_strategy (DistanceStrategy, optional) – Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
k (int, optional) – Number of Documents to return from search. Defaults to 4.
fetch_k (int, optional) – Number of Documents to fetch to pass to MMR algorithm. Defaults to 20.
lambda_mult (float, optional) – Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
index_query_options (Optional[QueryOptions], optional) – Index query option. Defaults to None.

Returns:

PGVectorStore

Return type:

PGVectorStore

delete(

ids: list | None = None,

**kwargs: Any,

) → bool | None[source]#

Delete records from the table.

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Parameters:

ids (list | None)
kwargs (Any)

Return type:

bool | None

drop_vector_index( index_name: str | None = None, ) → None[source]#

Drop the vector index.

Parameters:: index_name (str | None)
Return type:: None

classmethod from_documents(

documents: list[Document],

embedding: Embeddings,

engine: PGEngine,

table_name: str,

schema_name: str = 'public',

ids: list | None = None,

content_column: str = 'content',

embedding_column: str = 'embedding',

metadata_columns: list[str] | None = None,

ignore_metadata_columns: list[str] | None = None,

id_column: str = 'langchain_id',

metadata_json_column: str = 'langchain_metadata',

distance_strategy: DistanceStrategy = DistanceStrategy.COSINE_DISTANCE,

k: int = 4,

fetch_k: int = 20,

lambda_mult: float = 0.5,

index_query_options: QueryOptions | None = None,

**kwargs: Any,

) → PGVectorStore[source]#

Create an PGVectorStore instance from documents.

Parameters:

documents (list[Document]) – Documents to add to the vector store.
embedding (Embeddings) – Text embedding model to use.
engine (PGEngine) – Connection pool engine for managing connections to postgres database.
table_name (str) – Name of an existing table.
schema_name (str, optional) – Name of the database schema. Defaults to “public”.
ids (list | None) – (Optional[list]): List of IDs to add to table records. Defaults to None.
content_column (str, optional) – Column that represent a Document’s page_content. Defaults to “content”.
embedding_column (str, optional) – Column for embedding vectors. The embedding is generated from the document value. Defaults to “embedding”.
metadata_columns (list[str], optional) – Column(s) that represent a document’s metadata. Defaults to an empty list.
ignore_metadata_columns (Optional[list[str]], optional) – Column(s) to ignore in pre-existing tables for a document’s metadata. Can not be used with metadata_columns. Defaults to None.
id_column (str, optional) – Column that represents the Document’s id. Defaults to “langchain_id”.
metadata_json_column (str, optional) – Column to store metadata as JSON. Defaults to “langchain_metadata”.
distance_strategy (DistanceStrategy) – Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
k (int) – Number of Documents to return from search. Defaults to 4.
fetch_k (int) – Number of Documents to fetch to pass to MMR algorithm.
lambda_mult (float) – Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
index_query_options (QueryOptions) – Index query option.
kwargs (Any)

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Returns:

PGVectorStore

Return type:

PGVectorStore

classmethod from_texts(

texts: list[str],

embedding: Embeddings,

engine: PGEngine,

table_name: str,

schema_name: str = 'public',

metadatas: list[dict] | None = None,

ids: list | None = None,

content_column: str = 'content',

embedding_column: str = 'embedding',

metadata_columns: list[str] | None = None,

ignore_metadata_columns: list[str] | None = None,

id_column: str = 'langchain_id',

metadata_json_column: str = 'langchain_metadata',

distance_strategy: DistanceStrategy = DistanceStrategy.COSINE_DISTANCE,

k: int = 4,

fetch_k: int = 20,

lambda_mult: float = 0.5,

index_query_options: QueryOptions | None = None,

**kwargs: Any,

) → PGVectorStore[source]#

Create an PGVectorStore instance from texts.

Parameters:

texts (list[str]) – Texts to add to the vector store.
embedding (Embeddings) – Text embedding model to use.
engine (PGEngine) – Connection pool engine for managing connections to postgres database.
table_name (str) – Name of an existing table.
schema_name (str, optional) – Name of the database schema. Defaults to “public”.
metadatas (Optional[list[dict]], optional) – List of metadatas to add to table records. Defaults to None.
ids (list | None) – (Optional[list]): List of IDs to add to table records. Defaults to None.
content_column (str, optional) – Column that represent a Document’s page_content. Defaults to “content”.
embedding_column (str, optional) – Column for embedding vectors. The embedding is generated from the document value. Defaults to “embedding”.
metadata_columns (list[str], optional) – Column(s) that represent a document’s metadata. Defaults to empty list.
ignore_metadata_columns (Optional[list[str]], optional) – Column(s) to ignore in pre-existing tables for a document’s metadata. Can not be used with metadata_columns. Defaults to None.
id_column (str, optional) – Column that represents the Document’s id. Defaults to “langchain_id”.
metadata_json_column (str, optional) – Column to store metadata as JSON. Defaults to “langchain_metadata”.
distance_strategy (DistanceStrategy) – Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
k (int) – Number of Documents to return from search. Defaults to 4.
fetch_k (int) – Number of Documents to fetch to pass to MMR algorithm.
lambda_mult (float) – Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
index_query_options (QueryOptions) – Index query option.
kwargs (Any)

Raises:

InvalidTextRepresentationError <asyncpg.exceptions.InvalidTextRepresentationError> – if the ids data type does not match that of the id_column.

Returns:

PGVectorStore

Return type:

PGVectorStore

get_by_ids( ids: Sequence[str], ) → list[Document][source]#

Get documents by ids.

Parameters:: ids (Sequence[str])
Return type:: list[Document]

get_table_name() → str[source]#

Return type:: str

is_valid_index( index_name: str | None = None, ) → bool[source]#

Check if index exists in the table.

Parameters:: index_name (str | None)
Return type:: bool

max_marginal_relevance_search(

query: str,

k: int | None = None,

fetch_k: int | None = None,

lambda_mult: float | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected using the maximal marginal relevance.

Parameters:

query (str)
k (int | None)
fetch_k (int | None)
lambda_mult (float | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

max_marginal_relevance_search_by_vector(

embedding: list[float],

k: int | None = None,

fetch_k: int | None = None,

lambda_mult: float | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected using the maximal marginal relevance.

Parameters:

embedding (list[float])
k (int | None)
fetch_k (int | None)
lambda_mult (float | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

max_marginal_relevance_search_with_score_by_vector(

embedding: list[float],

k: int | None = None,

fetch_k: int | None = None,

lambda_mult: float | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[tuple[Document, float]][source]#

Return docs and distance scores selected using the maximal marginal relevance.

Parameters:

embedding (list[float])
k (int | None)
fetch_k (int | None)
lambda_mult (float | None)
filter (dict | None)
kwargs (Any)

Return type:

list[tuple[Document, float]]

reindex(index_name: str | None = None) → None[source]#

Re-index the vector store table.

Parameters:: index_name (str | None)
Return type:: None

search(

query: str,

search_type: str,

**kwargs: Any,

) → list[Document]#

Return docs most similar to query using a specified search type.

Parameters:

query (str) – Input text
search_type (str) – Type of search to perform. Can be “similarity”, “mmr”, or “similarity_score_threshold”.
**kwargs (Any) – Arguments to pass to the search method.

Returns:

List of Documents most similar to the query.

Raises:

ValueError – If search_type is not one of “similarity”, “mmr”, or “similarity_score_threshold”.

Return type:

list[Document]

similarity_search(

query: str,

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected by similarity search on query.

Parameters:

query (str)
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

similarity_search_by_vector(

embedding: list[float],

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[Document][source]#

Return docs selected by vector similarity search.

Parameters:

embedding (list[float])
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[Document]

similarity_search_with_relevance_scores(

query: str,

k: int = 4,

**kwargs: Any,

) → list[tuple[Document, float]]#

Return docs and relevance scores in the range [0, 1].

0 is dissimilar, 1 is most similar.

Parameters:

query (str) – Input text.
k (int) – Number of Documents to return. Defaults to 4.
**kwargs (Any) –
kwargs to be passed to similarity search. Should include: score_threshold: Optional, a floating point value between 0 to 1 to

filter the resulting set of retrieved docs.

Returns:

List of Tuples of (doc, similarity_score).

Return type:

list[tuple[Document, float]]

similarity_search_with_score(

query: str,

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[tuple[Document, float]][source]#

Return docs and distance scores selected by similarity search on query.

Parameters:

query (str)
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[tuple[Document, float]]

similarity_search_with_score_by_vector(

embedding: list[float],

k: int | None = None,

filter: dict | None = None,

**kwargs: Any,

) → list[tuple[Document, float]][source]#

Return docs and distance scores selected by similarity search on vector.

Parameters:

embedding (list[float])
k (int | None)
filter (dict | None)
kwargs (Any)

Return type:

list[tuple[Document, float]]