get_category_roles_stat

sparklightautoml.reader.guess_roles.get_category_roles_stat(train, subsample=100000, random_state=42)

Search for optimal processing of categorical values.

Categorical means defined by user or object types.

Parameters:

• train (SparkDataset) – Dataset.

• subsample (Union[float, int, None]) – size of subsample.

• random_state (int) – seed of random numbers generator.

Returns:

result.