constants Module

Defines names of feature types that are recognized for feature engineering in AutoML.

In typical use cases, you use FeatureType attributes for customizing featuration with the <xref:azureml.train.automl.automlconfig.AutoMLConfig> class and the featurization parameter.

Defines featurization modes that are used in AutoML.

In typical use cases, you specify featurization in the <xref:azureml.train.automl.automlconfig.AutoMLConfig> with the featurization parameter.

Constants relevant to the generation of the featurization run, along with default paths for artifacts to be transferred from the setup run.

Class storing nons-spaced languages' unicode ranges. Chinese is an example of a non-spaced language.

Names for prediction transform types

Constants for specifying what sort of prep run (e.g. setup only, setup before featurization, featurization only) we're doing before model selection.

Constants for Run History environment variable names.

A list of supported transformers with all customer name and factory method name.

Defines customer-facing names for transformers supported by AutoML.

Transformers are classified for use with Categorical data (e.g., CatImputer), DateTime data (e.g., DataTimeTransformer), Text data (e.g., TfIdf), or for Generic data types (e.g., Imputer).

Method names for transformers. These are Featurizer factory method names.

Defines transformer names for all transformers supported by AutoML, including those not exposed.

Defines mode names for feature and class balancing sweeping.

Class storing supported text dnn languages.

Names of neural models swept for text classification.

Transformer names with customer and factory method names.

Init TransformerName.

Transformer name mappings.

Defines parameters used by all transformers in AutoML.