weaviate.collections package

class weaviate.collections.Collection(connection: ConnectionV4, name: str, validate_arguments: bool, consistency_level: ConsistencyLevel | None = None, tenant: str | None = None, properties: Type[Properties] | None = None, references: Type[References] | None = None)[source]

Bases: Generic[Properties, References], _CollectionBase[Properties, References]

The collection class is the main entry point for interacting with a collection in Weaviate.

This class is returned by the client.collections.create and client.collections.get methods. It provides access to all the methods available to you when interacting with a collection in Weaviate.

You should not need to instantiate this class yourself but it may be useful to import this as a type when performing type hinting of functions that depend on a collection object.

Attributes:

aggregate: This namespace includes all the querying methods available to you when using Weaviate’s standard aggregation capabilities.
aggregate_group_by: This namespace includes all the aggregate methods available to you when using Weaviate’s aggregation group-by capabilities.
config: This namespace includes all the CRUD methods available to you when modifying the configuration of the collection in Weaviate.
data: This namespace includes all the CUD methods available to you when modifying the data of the collection in Weaviate.
generate: This namespace includes all the querying methods available to you when using Weaviate’s generative capabilities.
query_group_by: This namespace includes all the querying methods available to you when using Weaviate’s querying group-by capabilities.
query: This namespace includes all the querying methods available to you when using Weaviate’s standard query capabilities.
tenants: This namespace includes all the CRUD methods available to you when modifying the tenants of a multi-tenancy-enabled collection in Weaviate.

aggregate: This namespace includes all the querying methods available to you when using Weaviate’s standard aggregation capabilities.

backup: This namespace includes all the backup methods available to you when backing up a collection in Weaviate.

batch: This namespace contains all the functionality to upload data in batches to Weaviate for this specific collection.

config: This namespace includes all the CRUD methods available to you when modifying the configuration of the collection in Weaviate.

data: This namespace includes all the CUD methods available to you when modifying the data of the collection in Weaviate.

exists() → bool[source]: Check if the collection exists in Weaviate.

generate: This namespace includes all the querying methods available to you when using Weaviate’s generative capabilities.

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Sequence[str | QueryNested] | str | QueryNested | None = None, return_references: Literal[None] = None, after: str | UUID | None = None, cache_size: int | None = None) → _ObjectIterator[Properties, References][source]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Sequence[str | QueryNested] | str | QueryNested | None = None, return_references: Type[TReferences], after: str | UUID | None = None, cache_size: int | None = None) → _ObjectIterator[Properties, TReferences]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Type[TProperties], return_references: Literal[None] = None, after: str | UUID | None = None, cache_size: int | None = None) → _ObjectIterator[TProperties, References]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Type[TProperties], return_references: Sequence[_QueryReference | _QueryReferenceMultiTarget] | _QueryReference | _QueryReferenceMultiTarget, after: str | UUID | None = None, cache_size: int | None = None) → _ObjectIterator[TProperties, CrossReferences]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Type[TProperties], return_references: Type[TReferences], after: str | UUID | None = None, cache_size: int | None = None) → _ObjectIterator[TProperties, TReferences]

Use this method to return an iterator over the objects in the collection.

This iterator keeps a record of the last object that it returned to be used in each subsequent call to Weaviate. Once the collection is exhausted, the iterator exits.

If return_properties is not provided, all the properties of each object will be requested from Weaviate except for its vector as this is an expensive operation. Specify include_vector to request the vector back as well. In addition, if return_references=None then none of the references are returned. Use wvc.QueryReference to specify which references to return.

Arguments:

include_vector: Whether to include the vector in the metadata of the returned objects.
return_metadata: The metadata to return with each object.
return_properties: The properties to return with each object.
return_references: The references to return with each object.
after: The cursor to use to mark the initial starting point of the iterator in the collection.
cache_size: How many objects should be fetched in each request to Weaviate during the iteration. The default is 100.

Raises:

weaviate.exceptions.WeaviateGRPCQueryError:: If the request to the Weaviate server fails.

query: This namespace includes all the querying methods available to you when using Weaviate’s standard query capabilities.

shards() → List[Shard][source]

Get the statuses of all the shards of this collection.

Returns:

The list of shards belonging to this collection.

Raises

weaviate.WeaviateConnectionError: If the network connection to weaviate fails.
weaviate.UnexpectedStatusCodeError: If weaviate reports a none OK status.
weaviate.EmptyResponseError: If the response is empty.

tenants: This namespace includes all the CRUD methods available to you when modifying the tenants of a multi-tenancy-enabled collection in Weaviate.

with_consistency_level(consistency_level)

with_tenant(tenant)

class weaviate.collections.CollectionAsync(connection: ConnectionV4, name: str, validate_arguments: bool, consistency_level: ConsistencyLevel | None = None, tenant: str | None = None, properties: Type[Properties] | None = None, references: Type[References] | None = None)[source]

Bases: Generic[Properties, References], _CollectionBase[Properties, References]

The collection class is the main entry point for interacting with a collection in Weaviate.

This class is returned by the client.collections.create and client.collections.get methods. It provides access to all the methods available to you when interacting with a collection in Weaviate.

You should not need to instantiate this class yourself but it may be useful to import this as a type when performing type hinting of functions that depend on a collection object.

Attributes:

aggregate: This namespace includes all the querying methods available to you when using Weaviate’s standard aggregation capabilities.
aggregate_group_by: This namespace includes all the aggregate methods available to you when using Weaviate’s aggregation group-by capabilities.
config: This namespace includes all the CRUD methods available to you when modifying the configuration of the collection in Weaviate.
data: This namespace includes all the CUD methods available to you when modifying the data of the collection in Weaviate.
generate: This namespace includes all the querying methods available to you when using Weaviate’s generative capabilities.
query_group_by: This namespace includes all the querying methods available to you when using Weaviate’s querying group-by capabilities.
query: This namespace includes all the querying methods available to you when using Weaviate’s standard query capabilities.
tenants: This namespace includes all the CRUD methods available to you when modifying the tenants of a multi-tenancy-enabled collection in Weaviate.

aggregate: This namespace includes all the querying methods available to you when using Weaviate’s standard aggregation capabilities.

backup: This namespace includes all the backup methods available to you when backing up a collection in Weaviate.

config: This namespace includes all the CRUD methods available to you when modifying the configuration of the collection in Weaviate.

data: This namespace includes all the CUD methods available to you when modifying the data of the collection in Weaviate.

async exists() → bool[source]: Check if the collection exists in Weaviate.

generate: This namespace includes all the querying methods available to you when using Weaviate’s generative capabilities.

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Sequence[str | QueryNested] | str | QueryNested | None = None, return_references: Literal[None] = None, after: str | UUID | None = None, cache_size: int | None = None) → _ObjectAIterator[Properties, References][source]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Sequence[str | QueryNested] | str | QueryNested | None = None, return_references: Type[TReferences], after: str | UUID | None = None, cache_size: int | None = None) → _ObjectAIterator[Properties, TReferences]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Type[TProperties], return_references: Literal[None] = None, after: str | UUID | None = None, cache_size: int | None = None) → _ObjectAIterator[TProperties, References]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Type[TProperties], return_references: Sequence[_QueryReference | _QueryReferenceMultiTarget] | _QueryReference | _QueryReferenceMultiTarget, after: str | UUID | None = None, cache_size: int | None = None) → _ObjectAIterator[TProperties, CrossReferences]

iterator(include_vector: bool = False, return_metadata: List[Literal['creation_time', 'last_update_time', 'distance', 'certainty', 'score', 'explain_score', 'is_consistent']] | MetadataQuery | None = None, *, return_properties: Type[TProperties], return_references: Type[TReferences], after: str | UUID | None = None, cache_size: int | None = None) → _ObjectAIterator[TProperties, TReferences]

Use this method to return an iterator over the objects in the collection.

This iterator keeps a record of the last object that it returned to be used in each subsequent call to Weaviate. Once the collection is exhausted, the iterator exits.

If return_properties is not provided, all the properties of each object will be requested from Weaviate except for its vector as this is an expensive operation. Specify include_vector to request the vector back as well. In addition, if return_references=None then none of the references are returned. Use wvc.QueryReference to specify which references to return.

Arguments:

include_vector: Whether to include the vector in the metadata of the returned objects.
return_metadata: The metadata to return with each object.
return_properties: The properties to return with each object.
return_references: The references to return with each object.
after: The cursor to use to mark the initial starting point of the iterator in the collection.
cache_size: How many objects should be fetched in each request to Weaviate during the iteration. The default is 100.

Raises:

weaviate.exceptions.WeaviateGRPCQueryError:: If the request to the Weaviate server fails.

async length() → int[source]: Get the total number of objects in the collection.

query: This namespace includes all the querying methods available to you when using Weaviate’s standard query capabilities.

async shards() → List[Shard][source]

Get the statuses of all the shards of this collection.

Returns:

The list of shards belonging to this collection.

Raises

weaviate.WeaviateConnectionError: If the network connection to weaviate fails.
weaviate.UnexpectedStatusCodeError: If weaviate reports a none OK status.
weaviate.EmptyResponseError: If the response is empty.

tenants: This namespace includes all the CRUD methods available to you when modifying the tenants of a multi-tenancy-enabled collection in Weaviate.

async to_string() → str[source]: Return a string representation of the collection object.

with_consistency_level(consistency_level)

with_tenant(tenant)

Subpackages

Submodules

weaviate.collections.aggregate module

weaviate.collections.collection module

class weaviate.collections.collection.Collection(connection: ConnectionV4, name: str, validate_arguments: bool, consistency_level: ConsistencyLevel | None = None, tenant: str | None = None, properties: Type[Properties] | None = None, references: Type[References] | None = None)[source]

Bases: Generic[Properties, References], _CollectionBase[Properties, References]

The collection class is the main entry point for interacting with a collection in Weaviate.

This class is returned by the client.collections.create and client.collections.get methods. It provides access to all the methods available to you when interacting with a collection in Weaviate.

You should not need to instantiate this class yourself but it may be useful to import this as a type when performing type hinting of functions that depend on a collection object.

Attributes:

aggregate: This namespace includes all the querying methods available to you when using Weaviate’s standard aggregation capabilities.
aggregate_group_by: This namespace includes all the aggregate methods available to you when using Weaviate’s aggregation group-by capabilities.
config: This namespace includes all the CRUD methods available to you when modifying the configuration of the collection in Weaviate.
data: This namespace includes all the CUD methods available to you when modifying the data of the collection in Weaviate.
generate: This namespace includes all the querying methods available to you when using Weaviate’s generative capabilities.
query_group_by: This namespace includes all the querying methods available to you when using Weaviate’s querying group-by capabilities.
query: This namespace includes all the querying methods available to you when using Weaviate’s standard query capabilities.
tenants: This namespace includes all the CRUD methods available to you when modifying the tenants of a multi-tenancy-enabled collection in Weaviate.

aggregate: This namespace includes all the querying methods available to you when using Weaviate’s standard aggregation capabilities.

backup: This namespace includes all the backup methods available to you when backing up a collection in Weaviate.

batch: This namespace contains all the functionality to upload data in batches to Weaviate for this specific collection.

config: This namespace includes all the CRUD methods available to you when modifying the configuration of the collection in Weaviate.

data: This namespace includes all the CUD methods available to you when modifying the data of the collection in Weaviate.

exists() → bool[source]: Check if the collection exists in Weaviate.

generate: This namespace includes all the querying methods available to you when using Weaviate’s generative capabilities.