morpheus_llm.service.vdb.kinetica_vector_db_service.KineticaVectorDBResourceService

class KineticaVectorDBResourceService(name, schema, client)

Bases: VectorDBResourceService

Represents a service for managing resources in a Kinetica Vector Database.

Parameters:

namestr

Name of the Kinetica table. Must be an existing table

schemastr

Name of the Kinetica schema. Must be an existing schema

clientGPUdb instance

An instance of the GPUdb class for interaction with the Kinetica Vector Database.

Methods

count(**kwargs)	Returns number of rows/entities.
delete(expr, **kwargs)	Delete vectors from the Kinetica table using expressions.
delete_by_keys(keys, **kwargs)	Not supported by Kinetica.
describe(**kwargs)	Provides a description of the Kinetica table.
drop(**kwargs)	Drop a Kinetica table.
insert(data, **kwargs)	Insert data into the vector database.
insert_dataframe(df, **kwargs)	Insert a dataframe entires into the vector database.
query(query, **kwargs)	Query data in a table in the Kinetica database.
retrieve_by_keys(keys, **kwargs)	Retrieve the inserted vectors using their primary keys.
similarity_search(embeddings[, k])	Perform a similarity search within the Kinetica table.
similarity_search_by_vector(embedding, ...)	Return docs most similar to embedding vector.
update(data, **kwargs)	Update data in the Kinetica table.

similarity_search_with_score_by_vector

count(**kwargs)

Returns number of rows/entities.

Parameters:

**kwargsdict[str, typing.Any]

Not used.

Returns:

int

Returns number of records in the Kinetica table.

delete(expr, **kwargs)

Delete vectors from the Kinetica table using expressions.

Parameters:

exprstr

Delete expression.

**kwargsdict[str, typing.Any]

Extra keyword arguments specific to the vector database implementation.

Returns:

dict[str, typing.Any]

Returns result of the given keys that are deleted from the Kinetica table.

delete_by_keys(keys, **kwargs)

Not supported by Kinetica.

Parameters:

keysint | str | list

Primary keys to delete vectors.

**kwargsdict[str, typing.Any]

Extra keyword arguments specific to the vector database implementation.

Returns:

typing.Any

Returns vectors of the given keys that are delete from the resource.

describe(**kwargs)

Provides a description of the Kinetica table.

Parameters:

**kwargsdict[str, typing.Any]

Not used by Kinetica.

Returns:

dict

Returns response content as a dictionary.

drop(**kwargs)

Drop a Kinetica table.

This function allows you to delete/drop a Kinetica table.

Parameters:

**kwargsdict

Options as accepted by /clear/table API of Kinetica.

insert(data, **kwargs)

Insert data into the vector database.

Parameters:

datalist[list] | list[dict]

Data to be inserted into the Kinetica table.

**kwargsdict[str, typing.Any]

Extra keyword arguments specific to the vector database implementation.

Returns:

dict

Returns response content as a dictionary.

insert_dataframe(df, **kwargs)

Insert a dataframe entires into the vector database.

Parameters:

dfDataFrameType

Dataframe to be inserted into the Kinetica table.

**kwargsdict[str, typing.Any]

Not used by Kinetica.

Returns:

dict

Returns response content as a dictionary.

query(query, **kwargs)

Query data in a table in the Kinetica database.

This method performs a search operation in the specified table in the Kinetica database.

Parameters:

querystr, optional

The search query, which is an SQL query.

**kwargsdict

Additional keyword arguments for the search operation.

Returns:

typing.Any

The search result (a GPUdbSqlIterator object)

Raises:

GPUdbException

If an error occurs during the search operation.

retrieve_by_keys(keys, **kwargs)

Retrieve the inserted vectors using their primary keys.

Parameters:

keysint | str | list

Primary keys to get vectors for. Depending on pk_field type it can be int or str or a list of either.

**kwargsdict[str, typing.Any]

Additional keyword arguments for the retrieval operation.

Only valid keyword arguments are:

expression: [str] - The Kinetica expression to pass on to get/records/by/key

options: dict - The options dict accepted by get/records/by/key

Returns:

list[typing.Any]

Returns result rows of the given keys from the Kinetica table.

async similarity_search(

embeddings,

k=4,

**kwargs,

)

Perform a similarity search within the Kinetica table.

Parameters:

embeddingslist[list[float]]

Embeddings for which to perform the similarity search.

kint, optional

The number of nearest neighbors to return, by default 4.

**kwargsdict[str, typing.Any]

Extra keyword arguments specific to the vector database implementation.

Returns:

list[list[dict]]

Returns a list of ‘list of dictionaries’ representing the results of the similarity search.

similarity_search_by_vector(

embedding,

output_fields,

k=4,

kinetica_filter=None,

)

Return docs most similar to embedding vector.

Args:

embedding: Embedding to look up documents similar to. output_fields: The fields to return in the query output k: Number of Documents to return. Defaults to 4. kinetica_filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None.

Returns:

List of records most similar to the query vector.

update(data, **kwargs)

Update data in the Kinetica table.

Parameters:

datalist[typing.Any]

Data to be updated in the Kinetica table. This is npt used by Kinetica. The required parameters are all passed in a keyword arguments.

**kwargsdict[str, typing.Any]

Extra keyword arguments specific to Kinetica. The full explanation of each of these keyword arguments is available in the documentation of the API /update/records.

Allowed keyword arguments:

options: dict[str, str] - options

expressions: [list] - expression used to filter records for update

new_value_maps: [list[dict]] | [dict] -

records_to_insert: list[] -

records_to_insert_str: list[] -

Returns:

dict[str, typing.Any]

Returns result of the updated operation stats.