Ensemble

Ensemble is the entity representing results of the training process (done using the Task entity). It consists of a combination of one or more models, optimized for the dataset and model training configuration.

‘Ensemble’ API includes querying existing Ensembles (Get and List), but also the refit method, which allows users to maintain updated models. Future explainability features such as ROC curve, confusion matrix and other tools will be available as well.

class fireflyai.resources.ensemble.Ensemble[source]
classmethod delete(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Deletes a specific Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required, if fireflyai.authenticate() was run prior.
Returns:

“true” if deleted successfuly, raises FireflyClientError otherwise.
Return type:

FireflyResponse
classmethod edit_notes(id: int, notes: str, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Edits notes of the Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • notes (str) – New notes value.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

“submitted” if operation was successful, raises FireflyClientError otherwise.
Return type:

FireflyResponse
classmethod get(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Get information on a specific Ensemble.

Information includes the state of the Ensemble and other attributes.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Information about the Ensemble.
Return type:

FireflyResponse
classmethod get_ensemble_confusion_matrix(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets confusion matrix for Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Confusion matrix.
Return type:

FireflyResponse
classmethod get_ensemble_roc_curve(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets ROC curve data for Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

ROC curve data.
Return type:

FireflyResponse
classmethod get_ensemble_summary_report(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets summary report for Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Summary report.
Return type:

FireflyResponse
classmethod get_ensemble_test_prediction_sample(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets prediction samples for Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Prediction samples.
Return type:

FireflyResponse
classmethod get_model_architecture(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets architecture of the Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Architecture.
Return type:

FireflyResponse
classmethod get_model_presentation(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets presentation of the Ensemble.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Ensemble’s presentation.
Return type:

FireflyResponse
classmethod get_model_sensitivity_report(id: int, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

Gets sensitivity report for Ensemble.

Contains each feature’s sensitivity score for missing values and feature values.

Parameters:
  • id (int) – Ensemble ID.
  • api_key (Optional[str]) – Explicit api_key, not required if fireflyai.authenticate was run prior.
Returns:

Score for each feature in every sensitivity test.
Return type:

FireflyResponse
classmethod list(search_term: str = None, page: int = None, page_size: int = None, sort: Dict[KT, VT] = None, filter_: Dict[KT, VT] = None, api_key: str = None) → fireflyai.firefly_response.FireflyResponse[source]

List the existing Ensembles - supports filtering, sorting and pagination.

Parameters:
  • search_term (Optional[str]) – Return only records that contain the search_term in any field.
  • page (Optional[int]) – For pagination, which page to return.
  • page_size (Optional[int]) – For pagination, how many records will appear in a single page.
  • sort (Optional[Dict[str, Union[str, int]]]) – Dictionary of rules to sort the results by.
  • filter (Optional[Dict[str, Union[str, int]]]) – Dictionary of rules to filter the results by.
  • api_key (Optional[str]) – Explicit api_key, not required, if fireflyai.authenticate() was run prior.
Returns:

Ensembles are represented as nested dictionaries under hits.
Return type:

FireflyResponse