APIs

Top-level Classes

OpenMLBenchmarkSuite(suite_id, alias, name, …)	An OpenMLBenchmarkSuite represents the OpenML concept of a suite (a collection of tasks).
OpenMLClassificationTask(task_type_id, …)	OpenML Classification object.
OpenMLClusteringTask(task_type_id, …[, …])	OpenML Clustering object.
OpenMLDataFeature(index, name, data_type, …)	Data Feature (a.k.a.
OpenMLDataset(name, description[, format, …])	Dataset object.
OpenMLEvaluation(run_id, task_id, setup_id, …)	Contains all meta-information about a run / evaluation combination, according to the evaluation/list function
OpenMLFlow(name, description, model, …[, …])	OpenML Flow.
OpenMLLearningCurveTask(task_type_id, …[, …])	OpenML Learning Curve object.
OpenMLParameter(input_id, flow_id, …)	Parameter object (used in setup).
OpenMLRegressionTask(task_type_id, …[, …])	OpenML Regression object.
OpenMLRun(task_id, flow_id, dataset_id[, …])	OpenML Run: result of running a model on an openml dataset.
OpenMLSetup(setup_id, flow_id, parameters)	Setup object (a.k.a.
OpenMLSplit(name, description, split)	OpenML Split object.
OpenMLStudy(study_id, alias, …)	An OpenMLStudy represents the OpenML concept of a study (a collection of runs).
OpenMLSupervisedTask(task_type_id, …[, …])	OpenML Supervised Classification object.
OpenMLTask(task_id, task_type_id, task_type, …)	OpenML Task object.

Extensions

Extension	Defines the interface to connect machine learning libraries to OpenML-Python.
sklearn.SklearnExtension	Connect scikit-learn to OpenML-Python.
keras.KerasExtension	Connect Keras to OpenML-Python.
pytorch.PytorchExtension	Connect Pytorch to OpenML-Python.
onnx.OnnxExtension	Connect ONNX to OpenML-Python.
mxnet.MXNetExtension	Connect Apache MXNet to OpenML-Python.
mxnet.Config()	Represents the configuration of the OpenML MXNet Extension

get_extension_by_flow(flow, …)	Get an extension which can handle the given flow.
get_extension_by_model(model, …)	Get an extension which can handle the given flow.
register_extension(extension)	Register an extension.

Modules

openml.datasets: Dataset Functions

attributes_arff_from_df(df)	Describe attributes of the dataframe according to ARFF specification.
check_datasets_active(dataset_ids)	Check if the dataset ids provided are active.
create_dataset(name, description, creator, …)	Create a dataset.
get_dataset(dataset_id, str], download_data, …)	Download the OpenML dataset representation, optionally also download actual data file.
get_datasets(dataset_ids, int]], download_data)	Download datasets.
list_datasets(offset, NoneType] = None, …)	Return a list of all dataset which are on OpenML.
status_update(data_id, status)	Updates the status of a dataset to either ‘active’ or ‘deactivated’.

openml.evaluations: Evaluation Functions

list_evaluations(function, offset, …)

List all run-evaluation pairs matching all of the given filters.

openml.flows: Flow Functions

assert_flows_equal(flow1, flow2, …)	Check equality of two flows.
flow_exists(name, external_version)	Retrieves the flow id.
get_flow(flow_id, reinstantiate)	Download the OpenML flow for a given flow ID.
list_flows(offset, NoneType] = None, size, …)	Return a list of all flows which are on OpenML.

openml.runs: Run Functions

get_run(run_id, ignore_cache)	Gets run corresponding to run_id.
get_runs(run_ids)	Gets all runs in run_ids list.
get_run_trace(run_id)	Get the optimization trace object for a given run id.
initialize_model_from_run(run_id)	Initialized a model based on a run_id (i.e., using the exact same parameter settings)
initialize_model_from_trace(run_id, repeat, …)	Initialize a model based on the parameters that were set by an optimization procedure (i.e., using the exact same parameter settings)
list_runs(offset, NoneType] = None, size, …)	List all runs matching all of the given filters.
run_model_on_task(model, task, …)	Run the model on the dataset defined by the task.
run_flow_on_task(flow, task, …)	Run the model provided by the flow on the dataset defined by task.
run_exists(task_id, setup_id)	Checks whether a task/setup combination is already present on the server.

openml.setups: Setup Functions

get_setup(setup_id)	Downloads the setup (configuration) description from OpenML
initialize_model(setup_id)	Initialized a model based on a setup_id (i.e., using the exact same parameter settings)
list_setups(offset, NoneType] = None, size, …)	List all setups matching all of the given filters.
setup_exists(flow)	Checks whether a hyperparameter configuration already exists on the server.

openml.study: Study Functions

attach_to_study(study_id, run_ids)	Attaches a set of runs to a study.
attach_to_suite(suite_id, task_ids)	Attaches a set of tasks to a benchmarking suite.
create_benchmark_suite(name, description, …)	Creates an OpenML benchmark suite (collection of entity types, where the tasks are the linked entity)
create_study(name, description, run_ids, …)	Creates an OpenML study (collection of data, tasks, flows, setups and run), where the runs are the main entity (collection consists of runs and all entities (flows, tasks, etc) that are related to these runs)
delete_study(study_id)	Deletes a study from the OpenML server.
delete_suite(suite_id)	Deletes a study from the OpenML server.
detach_from_study(study_id, run_ids)	Detaches a set of run ids from a study.
detach_from_suite(suite_id, task_ids)	Detaches a set of task ids from a suite.
get_study(study_id, str], …)	Retrieves all relevant information of an OpenML study from the server.
get_suite(suite_id, str])	Retrieves all relevant information of an OpenML benchmarking suite from the server.
list_studies(offset, NoneType] = None, size, …)	Return a list of all studies which are on OpenML.
list_suites(offset, NoneType] = None, size, …)	Return a list of all suites which are on OpenML.
update_study_status(study_id, status)	Updates the status of a study to either ‘active’ or ‘deactivated’.
update_suite_status(suite_id, status)	Updates the status of a study to either ‘active’ or ‘deactivated’.

openml.tasks: Task Functions

get_task(task_id, download_data)	Download OpenML task for a given task ID.
get_tasks(task_ids[, download_data])	Download tasks.
list_tasks(task_type_id, NoneType] = None, …)	Return a number of tasks having the given tag and task_type_id Parameters ———- Filter task_type_id is separated from the other filters because it is used as task_type_id in the task description, but it is named type when used as a filter in list tasks call.