weaviate.classification

Module for classifying objects within Weaviate.

class weaviate.classification.Classification(connection: Connection)

Bases: object

Classification class used to schedule and/or check the status of a classification process of Weaviate objects.

Initialize a Classification class instance.

Parameters:: connection (weaviate.connect.Connection) – Connection object to an active and running Weaviate instance.

get(classification_uuid: str) → dict

Polls the current state of the given classification.

Parameters:

classification_uuid (str) – Identifier of the classification.

Returns:

A dict containing the Weaviate answer.

Return type:

dict

Raises:

ValueError – If not a proper uuid.
requests.ConnectionError – If the network connection to weaviate fails.
weaviate.UnexpectedStatusCodeException – If weaviate reports a none OK status.

is_complete(classification_uuid: str) → bool

Checks if a started classification job has completed.

Parameters:: classification_uuid (str) – Identifier of the classification.
Returns:: True if given classification has finished, False otherwise.
Return type:: bool

is_failed(classification_uuid: str) → bool

Checks if a started classification job has failed.

Parameters:: classification_uuid (str) – Identifier of the classification.
Returns:: True if the classification failed, False otherwise.
Return type:: bool

is_running(classification_uuid: str) → bool

Checks if a started classification job is running.

Parameters:: classification_uuid (str) – Identifier of the classification.
Returns:: True if the classification is running, False otherwise.
Return type:: bool

schedule() → ConfigBuilder

Schedule a Classification of the Objects within Weaviate.

Returns:: A ConfigBuilder that should be configured to the desired classification task
Return type:: weaviate.classification.config_builder.ConfigBuilder

class weaviate.classification.ConfigBuilder(connection: Connection, classification: Classification)

Bases: object

ConfigBuild class that is used to configure a classification process.

Initialize a ConfigBuilder class instance.

Parameters:

connection (weaviate.connect.Connection) – Connection object to an active and running weaviate instance.
classification (weaviate.classification.Classification) – Classification object to be configured using this ConfigBuilder instance.

do() → dict

Start the classification.

Returns:: Classification result.
Return type:: dict

with_based_on_properties(based_on_properties: list) → ConfigBuilder

Set properties to build the classification on.

Parameters:: based_on_properties (list) – A list of properties to classify on.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_class_name(class_name: str) → ConfigBuilder

What Object type to classify.

Parameters:: class_name (str) – Name of the class to be classified.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_classify_properties(classify_properties: list) → ConfigBuilder

Set the classify properties.

Parameters:: classify_properties (list) – A list of properties to classify.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_k(k: int) → ConfigBuilder

Set k number for the kNN.

Parameters:: k (int) – Number of objects to use to make a classification guess. (For kNN)
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_settings(settings: dict) → ConfigBuilder

Set settings for the classification. NOTE if you are using ‘kNN’ the value ‘k’ can be set by this method or by ‘with_k’. This method keeps previously set ‘settings’.

Parameters:: settings (dict) – Additional settings to be set/overwritten.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_source_where_filter(where_filter: dict) → ConfigBuilder

Set Source ‘where’ Filter.

Parameters:: where_filter (dict) – Filter to use, as a dict.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_target_where_filter(where_filter: dict) → ConfigBuilder

Set Target ‘where’ Filter.

Parameters:: where_filter (dict) – Filter to use, as a dict.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_training_set_where_filter(where_filter: dict) → ConfigBuilder

Set Training set ‘where’ Filter.

Parameters:: where_filter (dict) – Filter to use, as a dict.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_type(classification_type: str) → ConfigBuilder

Set classification type.

Parameters:: classification_type (str) – Type of the desired classification.
Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder

with_wait_for_completion() → ConfigBuilder

Wait for completion.

Returns:: Updated ConfigBuilder.
Return type:: ConfigBuilder