AbstractSampler 类

所有取样器的基类。

继承

ABC

AbstractSampler

构造函数

AbstractSampler(seed: int, max_rows: int = 10000, is_constraint_driven: bool = True, task: str = 'classification', train_frac: Optional[float] = None, *args: Any, **kwargs: Any)

参数

seed

max_rows

默认值: 10000

is_constraint_driven

默认值: True

task

默认值: classification

train_frac

默认值: None

方法

sample

所有子类都应进行此实现。

sample

所有子类都应进行此实现。

abstract sample(X: Union[numpy.ndarray, pandas.core.frame.DataFrame, scipy.sparse.base.spmatrix, azureml.dataprep.api.dataflow.Dataflow], y: Union[numpy.ndarray, pandas.core.series.Series, pandas.core.arrays.categorical.Categorical, azureml.dataprep.api.dataflow.Dataflow], cols: Optional[Union[str, List[str]]] = None) -> Tuple[Union[numpy.ndarray, pandas.core.frame.DataFrame, scipy.sparse.base.spmatrix, azureml.dataprep.api.dataflow.Dataflow], Union[numpy.ndarray, pandas.core.series.Series, pandas.core.arrays.categorical.Categorical, azureml.dataprep.api.dataflow.Dataflow], azureml.automl.runtime.sampling.splitting_config.SplittingConfig]

参数

X

必需

y

必需

cols

默认值: None