DRConfig#

class datarobotx.DRConfig(Data=None, Target=None, Featurization=None, Partitioning=None, Modeling=None, Metadata=None)[source]#

DataRobot configuration

Parameters that default to ‘None’ (or are omitted by the user) are overridden to server-side defaults at runtime. Consult the DataRobot REST API and GUI documentation for additional information on each parameter.

Parameters:

Data (dict or DataConfig) – Row and column selection configuration
Target (dict or TargetConfig) – Target configuration
Featurization (dict or FeaturesConfig) – Featurization configuration
Partitioning (dict or PartitioningConfig) – Partitioning configuration
Modeling (dict or ModelingConfig) – Modeling configuration
Metadata (dict or MetadataConfig) – DataRobot metadata and worker configuration

Examples

Direct configuration object construction

>>> config_1 = DRConfig()

Configuration object construction via abstraction get_params()

>>> model_1 = AutoMLModel(name='elated-varahamihira')
>>> config_2 = model.get_params()

Dictionary representation of configuration

>>> config_2
{'project_description': 'my_description', 'project_name': 'elated-varahamihira'}

Object specification through nested attributes (useful with autocomplete discovery)

>>> config_2.Metadata.project_description = 'my_description'

Configuration of abstractions using a config object, dictionary, or keyword arguments

>>> model_1.set_params(**config_2).get_params()
{'project_description': 'my_description', 'project_name': 'elated-varahamihira'}
>>> my_dict = {'project_description': 'my_new_description'}
>>> model_1.set_params(**my_dict).get_params()
{'project_description': 'my_new_description', 'project_name': 'elated-varahamihira'}
>>> model_1.set_params(project_name='my_new_name').get_params()
{'project_description': 'my_new_description', 'project_name': 'my_new_name'}

Construction of abstractions using a configuration object

>>> model_2 = AutoMLModel(**config_2)
>>> model_2.get_params()
{'project_description': 'my_description', 'project_name': 'elated-varahamihira'}

Attributes:

`Data`	Row and column selection configuration
`Featurization`	Featurization configuration
`Metadata`	DataRobot metadata and worker configuration
`Modeling`	Modeling configuration
`Partitioning`	Partitioning configuration
`Target`	Target configuration

Inherited methods:

`keys`()
`to_dict`()	Return configuration as a dict

property Data: DataConfig#

Row and column selection configuration

Notes

Data : dict or DataConfig

Return type:: DataConfig

property Featurization: FeaturesConfig#

Featurization configuration

Notes

Featurization : dict or FeaturesConfig

Return type:: FeaturesConfig

property Metadata: MetadataConfig#

DataRobot metadata and worker configuration

Notes

Metadata : dict or MetadataConfig

Return type:: MetadataConfig

property Modeling: ModelingConfig#

Modeling configuration

Notes

Modeling : dict or ModelingConfig

Return type:: ModelingConfig

property Partitioning: PartitioningConfig#

Partitioning configuration

Notes

Partitioning : dict or PartitioningConfig

Return type:: PartitioningConfig

property Target: TargetConfig#

Target configuration

Notes

Target : dict or TargetConfig

Return type:: TargetConfig

to_dict()[source]#: Return configuration as a dict